
!pip install numpy==1.20 --user

Requirement already satisfied: numpy==1.20 in c:\users\herik\appdata\roaming\python\python39\site-packages (1.20.0)


import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import matplotlib
import seaborn as sns
import re
import string
import nltk
import os
import folium
import mpl_toolkits
import plotly
import plotly.offline as pyoff
import plotly.graph_objs as go
import plotly.express as px
import chart_studio
import chart_studio.plotly as py
import holoviews as hv
import geoviews as gv
import datashader as ds

from warnings import simplefilter
from wordcloud import WordCloud
from nltk.corpus import stopwords
from nltk.stem import WordNetLemmatizer
from nltk.stem import PorterStemmer
from geopy.geocoders import Nominatim
from folium.plugins import HeatMap, FastMarkerCluster
from datetime import datetime, timedelta
from plotly.subplots import make_subplots
from colorcet import fire, rainbow, bgy, bjy, bkr, kb, kr
from datashader.colors import colormap_select, Greys9
from holoviews.streams import RangeXY
from holoviews.operation.datashader import datashade, dynspread, rasterize
from bokeh.io import push_notebook, show, output_notebook
from datashader.utils import lnglat_to_meters as webm
from datashader import transfer_functions as tf
from functools import partial
from datashader.utils import export_image
from IPython.display import HTML, display


os.environ['PROJ_LIB'] = '/Users/herik/anaconda3/Library/share/proj'


# Versões dos pacotes usados neste jupyter notebook
%reload_ext watermark
%watermark -a "Herikc Brecher" --iversions

Author: Herikc Brecher

folium      : 0.12.1.post1
numpy       : 1.20.0
seaborn     : 0.11.2
re          : 2.2.1
chart_studio: 1.1.0
matplotlib  : 3.5.1
geoviews    : 1.9.5
datashader  : 0.13.0
holoviews   : 1.14.8
pandas      : 1.4.2
plotly      : 5.6.0
nltk        : 3.7


simplefilter(action='ignore', category=FutureWarning)
%matplotlib inline
pyoff.init_notebook_mode()
sns.set_theme()
display(HTML("<style>.container { width:100% !important; }</style>"))


# Ambiente para Analise Geoespacial
output_notebook()
hv.extension('bokeh')

%opts Overlay [width=800 height=600 toolbar='above' xaxis=None yaxis=None]
%opts QuadMesh [tools=['hover'] colorbar=True] (alpha=0 hover_alpha=0.2)

T = 0.05
PX = 1


# Seed de Aleatoriedade
seed_ = 194
np.random.seed(seed_)


read_file = True

if read_file:
    try:
        with open('api.txt') as f:
            lines = f.readlines()

        # Ler usuario
        username = lines[0].split(':')[1].split('\n')[0]

        # Ler api key
        api_key = lines[1].split(':')[1]
    except:
        read_file = False


upload_to_dash_desejado = False

if read_file:
    # Configurando key para upar no dash do plotly
    chart_studio.tools.set_credentials_file(username=username, api_key=api_key)

    # Configurando Privacidade
    chart_studio.tools.set_config_file(world_readable = True, sharing = 'public')

    # Variavel para upar no dash
    upload_to_dash = upload_to_dash_desejado
else:
    upload_to_dash = False


# Carregamento de todos datasets
dtCustomers = pd.read_csv('../data/olist_customers_dataset.csv', encoding = 'utf8', dtype={'customer_zip_code_prefix': str})
dtGeolocation = pd.read_csv('../data/olist_geolocation_dataset.csv', encoding = 'utf8', dtype={'geolocation_zip_code_prefix': str})
dtOrderItems = pd.read_csv('../data/olist_order_items_dataset.csv', encoding = 'utf8')
dtOrderPayments = pd.read_csv('../data/olist_order_payments_dataset.csv', encoding = 'utf8')
dtOrderReviews = pd.read_csv('../data/olist_order_reviews_dataset.csv', encoding = 'utf8')
dtOrders = pd.read_csv('../data/olist_orders_dataset.csv', encoding = 'utf8')
dtProducts = pd.read_csv('../data/olist_products_dataset.csv', encoding = 'utf8')
dtSellers = pd.read_csv('../data/olist_sellers_dataset.csv', encoding = 'utf8')
dtCategory = pd.read_csv('../data/product_category_name_translation.csv', encoding = 'utf8')


regioes = {
    'AC': 'Norte',
    'AL': 'Nordeste',
    'AP': 'Norte',
    'AM': 'Norte',
    'BA': 'Nordeste',
    'CE': 'Nordeste',
    'DF': 'CentroOeste',
    'ES': 'Sudeste',
    'GO': 'CentroOeste',
    'MA': 'Nordeste',
    'MT': 'CentroOeste',
    'MS': 'CentroOeste',
    'MG': 'Sudeste',
    'PA': 'Norte',
    'PB': 'Nordeste',
    'PR': 'Sul',
    'PE': 'Nordeste',
    'PI': 'Nordeste',
    'RJ': 'Sudeste',
    'RN': 'Nordeste',
    'RS': 'Sul',
    'RO': 'Norte',
    'RR': 'Norte',
    'SC': 'Sul',
    'SP': 'Sudeste',
    'SE': 'Nordeste',
    'TO': 'Norte'
}


dtCustomers['regiao'] =  dtCustomers['customer_state'].copy()


dtCustomers = dtCustomers.replace({'regiao': regioes})


dtCustomers.head()


dtSellers['regiao'] =  dtSellers['seller_state'].copy()


dtSellers = dtCustomers.replace({'regiao': regioes})


dtSellers.head()


dtGeolocation['regiao'] =  dtGeolocation['geolocation_state'].copy()


dtGeolocation = dtGeolocation.replace({'regiao': regioes})


dtGeolocation.head()


# Dataset de pedidos
dtOrders.head()


dtOrders.dtypes

order_id                         object
customer_id                      object
order_status                     object
order_purchase_timestamp         object
order_approved_at                object
order_delivered_carrier_date     object
order_delivered_customer_date    object
order_estimated_delivery_date    object
dtype: object


# Colunas do tipo data
dateColumns = ['order_purchase_timestamp', 'order_approved_at', 'order_delivered_carrier_date',\
               'order_delivered_customer_date', 'order_estimated_delivery_date']

# Dataset de analise temporal
dtOrdersAjustado = dtOrders.copy()


# Convertendo columas de data para date
for col in dateColumns:
    dtOrdersAjustado[col] = pd.to_datetime(dtOrdersAjustado[col], format = '%Y-%m-%d %H:%M:%S')


dtOrdersAjustado.head()


dtOrdersAjustado.dtypes

order_id                                 object
customer_id                              object
order_status                             object
order_purchase_timestamp         datetime64[ns]
order_approved_at                datetime64[ns]
order_delivered_carrier_date     datetime64[ns]
order_delivered_customer_date    datetime64[ns]
order_estimated_delivery_date    datetime64[ns]
dtype: object


# Verificando tipos unicos
print(dtOrdersAjustado.nunique())

order_id                         99441
customer_id                      99441
order_status                         8
order_purchase_timestamp         98875
order_approved_at                90733
order_delivered_carrier_date     81018
order_delivered_customer_date    95664
order_estimated_delivery_date      459
dtype: int64


# Verificando columas com null
print(dtOrdersAjustado.isna().sum())

order_id                            0
customer_id                         0
order_status                        0
order_purchase_timestamp            0
order_approved_at                 160
order_delivered_carrier_date     1783
order_delivered_customer_date    2965
order_estimated_delivery_date       0
dtype: int64


# Verificando valores NA
dtOrdersAjustado[dtOrdersAjustado.isna().any(axis = 1)]


# Dropando valores NA
dtOrdersAjustado = dtOrdersAjustado.dropna()


# Verificando se possuem valores duplicados
print(sum(dtOrdersAjustado.duplicated()))

0


dtOrdersAjustado.head()


# Dataframe all
dtAll = dtOrders.merge(dtCustomers, on="customer_id").merge(dtOrderItems, on="order_id").merge(dtProducts, on="product_id").merge(dtCategory, on="product_category_name").merge(dtOrderPayments, on="order_id").merge(dtOrderReviews, on="order_id")


dtOrdersMesAno = dtOrdersAjustado.copy()

# Mantendo somente Ano e Mes
for col in dateColumns:
    dtOrdersMesAno[col] = pd.to_datetime(dtOrdersMesAno[col], format = '%Y-%m-%d %H:%M:%S').dt.to_period('M')


dtOrdersMesAno.head()


dtFiltrado = dtOrdersMesAno.groupby('order_approved_at').count()['order_id'].to_frame().reset_index()


dtFiltrado['order_approved_at'] = dtFiltrado['order_approved_at'].astype(str)


# Plot

# Definição dos dados no plot
plot_data = [go.Scatter(x = dtFiltrado['order_approved_at'],
y = dtFiltrado['order_id'])]

# Layout
plot_layout = go.Layout(xaxis = {"type": "category", "title": "Periodo"},
                        yaxis = {"title": "Vendas"},
                        title = 'Evolução das Vendas Mensais')

# Plot da figura
fig = go.Figure(data = plot_data, layout = plot_layout)

fig.add_annotation(
        x = 13,
        y = 7149,
        xref="x",
        yref="y",
        text="Máximo de Vendas",
        showarrow = True,
        font=dict(
            family="Courier New, monospace",
            size=16,
            color="#636363"
            ),
        align="center",
        arrowhead=2,
        arrowsize=1,
        arrowwidth=2,
        arrowcolor="#636363",
        ax=20,
        ay=-30,
        bordercolor="#636363",
        borderwidth=1.5,
        borderpad=4,
        bgcolor="#fff",
        opacity=1
        )

pyoff.iplot(fig)

# Upando no dash
if upload_to_dash:
    py.plot(fig, filename = 'evolucao_vendas_mensais', auto_open = False)


dtOrderPrice = dtOrderItems.groupby('order_id').agg({'price': sum}).reset_index()


dtOrderPrice.head()


dtFaturamentoMesAno = pd.merge(dtOrdersMesAno, dtOrderPrice, on = 'order_id')


dtFaturamentoMesAno.head()


dtFiltrado = dtFaturamentoMesAno.groupby('order_approved_at').agg({'price': sum}).reset_index()


dtFiltrado['order_approved_at'] = dtFiltrado['order_approved_at'].astype(str)


dtFiltrado.head()


# Plot

# Definição dos dados no plot
plot_data = [go.Scatter(x = dtFiltrado['order_approved_at'], 
                        y = dtFiltrado['price'])]

# Layout
plot_layout = go.Layout(xaxis = {"type": "category", 'title': 'Periodo'},
                        yaxis = {"title": 'Faturamento'},
                        title = 'Evolução Faturamento Mensal')

# Plot da figura
fig = go.Figure(data = plot_data, layout = plot_layout)


fig.add_annotation(
        x = 19,
        y = 999867,
        xref="x",
        yref="y",
        text="Máximo de Faturamento",
        showarrow = True,
        font=dict(
            family="Courier New, monospace",
            size=16,
            color="#000"
            ),
        align="center",
        arrowhead=2,
        arrowsize=1,
        arrowwidth=2,
        arrowcolor="#636363",
        ax=20,
        ay=-30,
        bordercolor="#636363",
        borderwidth=1.5,
        borderpad=4,
        bgcolor="#fff",
        opacity=1
        )

pyoff.iplot(fig)

# Upando no dash
if upload_to_dash:
    py.plot(fig, filename = 'evolucao_faturamento_mensal', auto_open = False)


months = {
            '01': 'Janeiro',
            '02': 'Fevereiro',
            '03': 'Março',
            '04': 'Abril',
            '05': 'Maio',
            '06': 'Junho',
            '07': 'Julho',
            '08': 'Agosto',
            '09': 'Setembro',
            '10': 'Outubro',
            '11': 'Novembro',
            '12': 'Dezembro'        
        }


dtFiltrado2017 = dtFiltrado.copy()
dtFiltrado2017 = dtFiltrado2017.query("'2017-01' <= order_approved_at <= '2017-08'")


dtFiltrado2017['order_approved_at'] = dtFiltrado2017['order_approved_at'].apply(lambda x: x[5:])


dtFiltrado2017 = dtFiltrado2017.replace({'order_approved_at': months})


dtFiltrado2017.columns = ['Mes-2017', 'Faturamento-2017']


dtFiltrado2017 = dtFiltrado2017.reset_index()
dtFiltrado2017 = dtFiltrado2017.drop('index', axis = 1)


dtFiltrado2017.head()


dtFiltrado2018 = dtFiltrado.copy()
dtFiltrado2018 = dtFiltrado2018.query("'2018-01' <= order_approved_at <= '2018-08'")


dtFiltrado2018['order_approved_at'] = dtFiltrado2018['order_approved_at'].apply(lambda x: x[5:])


dtFiltrado2018 = dtFiltrado2018.replace({'order_approved_at': months})


dtFiltrado2018.columns = ['Mes-2018', 'Faturamento-2018']


dtFiltrado2018 = dtFiltrado2018.reset_index()
dtFiltrado2018 = dtFiltrado2018.drop('index', axis = 1)


dtFaturamento2017To2018 = pd.concat([dtFiltrado2017, dtFiltrado2018], axis = 1)


dtFaturamento2017To2018['crescimento-2017-para-2018'] = dtFaturamento2017To2018.apply( lambda x: str( round((x[3] / x[1]) * 100, 2)), axis = 1 )


dtFaturamento2017To2018['crescimento-2017-para-2018'] =  dtFaturamento2017To2018.apply(lambda x: '+' + x[4] + '%' if x[3] > x[1] else '-' + x[4] + '%', axis = 1 )


dtFaturamento2017To2018


# Plot

# Definição dos dados no plot
plot_data = [go.Bar(name = '2017', x = dtFaturamento2017To2018['Mes-2017'], y = dtFaturamento2017To2018['Faturamento-2017']),
             go.Bar(name = '2018', x = dtFaturamento2017To2018['Mes-2018'], y = dtFaturamento2017To2018['Faturamento-2018'], text = dtFaturamento2017To2018['crescimento-2017-para-2018'], textposition = 'outside')]

# Layout
plot_layout = go.Layout(xaxis = {'title': 'Periodo'},
                        yaxis = {"title": 'Faturamento'},
                        barmode = 'group',
                        title = 'Faturamento Mensal 2017 Vs 2018')

# Plot da figura
fig = go.Figure(data = plot_data, layout = plot_layout)

pyoff.iplot(fig)

# Upando no dash
if upload_to_dash:
    py.plot(fig, filename = 'faturamento_mensal_2017_vs_2018', auto_open = False)


# Calculando crescimento Mensal
dtFiltrado['MonthlyGrowth'] = dtFiltrado['price'].pct_change()


dtFiltrado.head()


# Plot

# Definição dos dados no plot (Iniciando em Fevereiro de 2017 para não destorcer os dados)
plot_data = [go.Scatter(x = dtFiltrado.query("order_approved_at > '2017-01'")['order_approved_at'],
                        y = dtFiltrado.query("order_approved_at > '2017-01'")['MonthlyGrowth'])]

# Layout
plot_layout = go.Layout(xaxis = {"type": "category", 'title': 'Periodo'},
                        yaxis = {"title": 'Crescimento em Porcentagem'},
                        title = 'Taxa Percentual de Crescimento Mensal')

# Plot da figura
fig = go.Figure(data = plot_data, layout = plot_layout)

fig.add_annotation(
        x = 0,
        y = 1.2,
        xref="x",
        yref="y",
        text="Crescimento Residual",
        showarrow = True,
        font=dict(
            family="Courier New, monospace",
            size=16,
            color="#000"
            ),
        align="center",
        arrowhead=2,
        arrowsize=1,
        arrowwidth=2,
        arrowcolor="#636363",
        ax=20,
        ay=-30,
        bordercolor="#636363",
        borderwidth=1.5,
        borderpad=4,
        bgcolor="#fff",
        opacity=1
        )

fig.add_annotation(
        x = 9,
        y = 0.49,
        xref="x",
        yref="y",
        text="Maior Taxa de Crescimento",
        showarrow = True,
        font=dict(
            family="Courier New, monospace",
            size=16,
            color="#000"
            ),
        align="center",
        arrowhead=2,
        arrowsize=1,
        arrowwidth=2,
        arrowcolor="#636363",
        ax=20,
        ay=-30,
        bordercolor="#636363",
        borderwidth=1.5,
        borderpad=4,
        bgcolor="#fff",
        opacity=1
        )

pyoff.iplot(fig) 

# Upando no dash
if upload_to_dash:
    py.plot(fig, filename = 'taxa_percentual_crescimento_brasil', auto_open = False)


plt.figure(figsize = (15, 7))
dtOrdersMesAno.groupby('order_approved_at').count()['order_id'].plot(kind = 'bar')

plt.xlabel('Mes / Ano')
plt.ylabel('Vendas')
plt.title('Vendas por Mes/Ano')

plt.show()


# Criando copia dataset ajustado
dtOrdersMesAnoDia = dtOrdersAjustado[['order_id', 'order_approved_at']].copy()


dtOrdersMesAnoDia['weekday'] = dtOrdersMesAnoDia['order_approved_at'].dt.dayofweek


dtOrdersMesAnoDia.head()


# Mantendo somente Ano, Mes e Dia
dtOrdersMesAnoDia['order_approved_at'] = pd.to_datetime(dtOrdersMesAnoDia['order_approved_at'],\
                                                        format = '%Y-%m-%d %H:%M:%S').dt.to_period('y')


dtOrdersMesAnoDia.head()


dtFiltrado = dtOrdersMesAnoDia.groupby('weekday').count()['order_id'].reset_index()


dtFiltrado['percent'] = dtFiltrado['order_id'].apply(lambda x: str( round(x / sum(dtFiltrado['order_id']) * 100, 2) ) + '%')


# Plot

# Definição dos dados no plot
labels = ['Domingo', 'Segunda', 'Terça', 'Quarta', 'Quinta', 'Sexta', 'Sabado']
plot_data = [go.Bar(x = labels, 
                    y = dtFiltrado['order_id'],
                    text = dtFiltrado['percent'], textposition = 'outside')] #, textfont=dict(color='#E58606')

# Layout
plot_layout = go.Layout(xaxis = {"type": "category", 'title': 'Dia da Semana'},
                        yaxis = {'title': 'Vendas'},
                        title = 'Total de Vendas por Dia da Semana')

# Plot da figura
fig = go.Figure(data = plot_data, layout = plot_layout)
pyoff.iplot(fig)

# Upando no dash
if upload_to_dash:
    py.plot(fig, filename = 'total_vendas_dia_semana', auto_open = False)


dtOrdersMesAnoDia2016 = dtOrdersMesAnoDia[dtOrdersMesAnoDia['order_approved_at'] == '2016'].groupby('weekday').\
count()['order_id']
dtOrdersMesAnoDia2017 = dtOrdersMesAnoDia[dtOrdersMesAnoDia['order_approved_at'] == '2017'].groupby('weekday').\
count()['order_id']
dtOrdersMesAnoDia2018 = dtOrdersMesAnoDia[dtOrdersMesAnoDia['order_approved_at'] == '2018'].groupby('weekday').\
count()['order_id']


dtOrdersMesAnoDiaAgrupado = dtOrdersMesAnoDia.groupby(['order_approved_at', 'weekday']).count().reset_index()


dtOrdersMesAnoDiaAgrupado.columns = ['year', 'weekday', 'orders']


dtOrdersMesAnoDiaAgrupado['year'] = dtOrdersMesAnoDiaAgrupado['year'].astype(str)


dtOrdersMesAnoDiaAgrupado


dtOrdersMesAnoDiaPivot = dtOrdersMesAnoDiaAgrupado.pivot(index = 'weekday', columns = 'year')['orders'].reset_index()


dtOrdersMesAnoDiaPivot['weekday'] = dtOrdersMesAnoDiaPivot['weekday'].astype(str)


labels_ = {'0': 'Domingo', '1': 'Segunda', '2': 'Terça', '3': 'Quarta', '4': 'Quinta', '5': 'Sexta', '6': 'Sabado'}


dtOrdersMesAnoDiaPivot = dtOrdersMesAnoDiaPivot.replace({'weekday': labels_})


dtOrdersMesAnoDiaPivot.head()


fig = px.bar(dtOrdersMesAnoDiaPivot,\
             x = "weekday", y = ['2016', '2017', '2018'], title = "Total de Vendas por Dia da Semana por Ano", labels = {'variable': 'Ano', 'value': 'Vendas', 'weekday': 'Dia da Semana'})
fig.show()

# Upando no dash
if upload_to_dash:
    py.plot(fig, filename = 'total_vendas_dia_semana_por_ano', auto_open = False)


# Criando copia para novo dataset
dtOrdersDiaMesAno = dtOrdersAjustado[['order_id', 'order_approved_at']].copy()

# Mantendo somente Ano, Mes e Dia
dtOrdersDiaMesAno['Data'] = \
    pd.to_datetime(dtOrdersDiaMesAno['order_approved_at'], format = '%Y-%m-%d %H:%M:%S').dt.to_period('d')


# Criandop coluna de Ano
dtOrdersDiaMesAno['Ano'] = \
    pd.to_datetime(dtOrdersDiaMesAno['order_approved_at'], format = '%Y-%m-%d %H:%M:%S').dt.to_period('y')


# Criando coluna de weekday
dtOrdersDiaMesAno['weekday'] = dtOrdersDiaMesAno['order_approved_at'].dt.dayofweek


dtOrdersDiaMesAno.head()


def cria_media_semanal_por_ano(data, ano):
    dtAnoDia = data[data['Ano'] == ano].groupby('weekday').nunique()
    dtAnoDia['mean_orders'] = dtAnoDia.apply(lambda x: int(x['order_id'] / x['Data']), axis = 1)
    dtAnoDia = dtAnoDia['mean_orders'].reset_index()
    dtAnoDia['year'] = ano
    dtAnoDia = dtAnoDia.pivot(index = 'weekday', columns = 'year')['mean_orders'].reset_index()
    
    return dtAnoDia


# Calcula media de vendas por dia da semana do ano de 2016
dtOrdersDiaMesAno2016 = cria_media_semanal_por_ano(dtOrdersDiaMesAno, '2016')


# Calcula media de vendas por dia da semana do ano de 2017
dtOrdersDiaMesAno2017 = cria_media_semanal_por_ano(dtOrdersDiaMesAno, '2017')


# Calcula media de vendas por dia da semana  do ano de 2018
dtOrdersDiaMesAno2018 = cria_media_semanal_por_ano(dtOrdersDiaMesAno, '2018')


dtOrdersDiaMesAno2016


# Concatenando coluna de cada ano
dtOrdersDiaMesAnoPivot = dtOrdersDiaMesAno2016.copy()
dtOrdersDiaMesAnoPivot['2017'] = dtOrdersDiaMesAno2017['2017']
dtOrdersDiaMesAnoPivot['2018'] = dtOrdersDiaMesAno2018['2018']


dtOrdersDiaMesAnoPivot['weekday'] = dtOrdersDiaMesAnoPivot['weekday'].astype(str)


dtOrdersDiaMesAnoPivot = dtOrdersDiaMesAnoPivot.replace({'weekday': labels_})


dtOrdersDiaMesAnoPivot


fig = px.bar(dtOrdersDiaMesAnoPivot,\
             x = "weekday", y = ['2016', '2017', '2018'], title = "Média de Vendas por Dia da Semana por Ano", labels = {'variable': 'Ano', 'value': 'Vendas', 'weekday': 'Dia da Semana'})
fig.show()

# Upando no dash
if upload_to_dash:
    py.plot(fig, filename = 'media_vendas_dia_semana_ano', auto_open = False)


dtOrdersAjustado.head()


dtOrdersDelivered = dtOrdersAjustado[dtOrdersAjustado['order_status'] == 'delivered'].copy()


dtOrdersDelivered.dtypes

order_id                                 object
customer_id                              object
order_status                             object
order_purchase_timestamp         datetime64[ns]
order_approved_at                datetime64[ns]
order_delivered_carrier_date     datetime64[ns]
order_delivered_customer_date    datetime64[ns]
order_estimated_delivery_date    datetime64[ns]
dtype: object


def parteDia(hour):
    value = 'Madrugada'
    
    if 6 <= hour <= 11:
        value = 'Manhã'
    elif 12 <= hour <= 17:
        value = 'Tarde'
    elif 18 <= hour <= 23:
        value = 'Noite'
        
    return value


dtOrdersDelivered['purchase_part_day'] = dtOrdersDelivered['order_purchase_timestamp'].apply(lambda x: parteDia(x.hour))


dtOrdersDelivered.head()


dtFiltrado = dtOrdersDelivered.groupby('purchase_part_day').count()['order_id'].reset_index()


dtFiltrado['percent'] = dtFiltrado['order_id'].apply(lambda x: str( round(x / sum(dtFiltrado['order_id']) * 100, 2) ) + '%')


dtFiltrado.head()


# Plot

# Definição dos dados no plot
plot_data = [go.Bar(x = dtFiltrado['purchase_part_day'],
                    y = dtFiltrado['order_id'],
                    text = dtFiltrado['percent'], textposition = 'outside')]

# Layout
plot_layout = go.Layout(xaxis = {"type": "category", 'title': 'Periodo do Dia'}, 
                        yaxis = {'title': 'Vendas'}, 
                        title = 'Total de Vendas por Período do Dia')

# Plot da figura
fig = go.Figure(data = plot_data, layout = plot_layout)
pyoff.iplot(fig)

# Upando no dash
if upload_to_dash:
    py.plot(fig, filename = 'total_vendas_periodo_dia', auto_open = False)


dtOrderPayments.head()


dtOrderPayments.groupby('payment_type').count()


dtOrderPayments = dtOrderPayments[dtOrderPayments['payment_type'] != 'not_defined']


dtOrderPayments.groupby('payment_type').count()


# Definição dos dados no plot
labels = ['Boleto', 'Cartão de Crédito', 'Cartão de Débito', 'Voucher']
values = dtOrderPayments.groupby('payment_type').count()['order_id']

plot_data = [go.Pie(labels = labels, values = values, hole = .3, pull = [0, .1, 0, 0], name = 'Pagamentos')]

# Layout
plot_layout = go.Layout(title={'text': 'Métodos de Pagamentos', 'y':0.85, 'x':0.51, 'xanchor': 'center', 'yanchor': 'top'})

# Plot da figura
fig = go.Figure(data = plot_data, layout = plot_layout)

fig.update_layout(legend_title_text='Meios de Pagamento', legend=dict(
    yanchor="top",
    y = 0.80,
    xanchor="right",
    x = 0.85
))

pyoff.iplot(fig)

# Upando no dash
if upload_to_dash:
    py.plot(fig, filename = 'metodos_pagamentos', auto_open = False)


dtCustomers.head()


salesState = dtCustomers.groupby('customer_state')['customer_id'].count()


salesState = salesState.sort_values(ascending = False).reset_index()


salesState['percent'] = salesState['customer_id'].apply(lambda x: str( round(x / sum(salesState['customer_id']) * 100, 2) ) + '%')


salesState.head()


# Plot

# Definição dos dados no plot
plot_data = [go.Bar(x = salesState['customer_state'], 
                    y = salesState['customer_id'],
                    text = salesState['percent'], textposition = 'outside')]

# Layout
plot_layout = go.Layout(xaxis = {"type": "category", 'title': 'Estados'},
                        yaxis = {'title': 'Vendas'},
                        title = 'Total de Vendas por Estado no Brasil')

# Plot da figura
fig = go.Figure(data = plot_data, layout = plot_layout)
pyoff.iplot(fig)

# Upando no dash
if upload_to_dash:
    py.plot(fig, filename = 'total_vendas_estado_brasil', auto_open = False)


salesRegiao = dtCustomers.groupby('regiao')['customer_id'].count().reset_index()


salesRegiao['percent'] = salesRegiao['customer_id'].apply(lambda x: str( round(x / sum(salesRegiao['customer_id']) * 100, 2) ) + '%')


salesRegiao['percent_number'] = salesRegiao['customer_id'].apply(lambda x: round(x / sum(salesRegiao['customer_id']) * 100, 2) )


salesRegiao = salesRegiao.sort_values('percent_number', ascending  = False)


salesRegiao


# Plot

# Definição dos dados no plot
plot_data = [go.Bar(x = salesRegiao['regiao'], 
                    y = salesRegiao['customer_id'],
                    text = salesRegiao['percent'], textposition = 'outside')]

# Layout
plot_layout = go.Layout(xaxis = {"type": "category", 'title': 'Regiões'},
                        yaxis = {'title': 'Vendas'},
                        title = 'Total de Vendas por Região no Brasil')

# Plot da figura
fig = go.Figure(data = plot_data, layout = plot_layout)
pyoff.iplot(fig)

# Upando no dash
if upload_to_dash:
    py.plot(fig, filename = 'total_vendas_regiao_brasil', auto_open = False)


salesCidade = dtCustomers.groupby('customer_city')['customer_id'].count()


salesCidade10 = salesCidade.sort_values(ascending = False).reset_index().head(10)


salesCidade = salesCidade.reset_index()


salesCidade10['percent'] = salesCidade10['customer_id'].apply(lambda x: str( round(x / sum(salesCidade['customer_id']) * 100, 2) ) + '%')


salesCidade10 = salesCidade10.sort_values(ascending = True, by = 'customer_id')


# Plot

# Definição dos dados no plot
plot_data = [go.Bar(y = salesCidade10['customer_city'], 
                    x = salesCidade10['customer_id'], orientation = 'h',
                    text = salesCidade10['percent'], textposition = 'outside')]

# Layout
plot_layout = go.Layout(yaxis = {"type": "category", 'title': 'Cidades'},
                        xaxis = {'title': 'Vendas'},
                        title = 'Top 10 Cidades com Maior Numero de Vendas')

# Plot da figura
fig = go.Figure(data = plot_data, layout = plot_layout)
pyoff.iplot(fig)

# Upando no dash
if upload_to_dash:
    py.plot(fig, filename = 'top_10_cidades_vendas', auto_open = False)


topRevenueByCity = dtAll.groupby("customer_city")["payment_value"].sum().reset_index().sort_values("payment_value", ascending=False)
topRevenueByCity["% of Total Payments"] = (topRevenueByCity["payment_value"] / topRevenueByCity["payment_value"].sum()) * 100
topRevenueByCity["% Cumulativa dos Pagamentos"] = topRevenueByCity["% of Total Payments"].cumsum()


topRevenueByCity.head()


sns.set_style("ticks")

plt.figure(figsize=(12, 8))

ax = sns.lineplot(x=range(1,len(topRevenueByCity)+1), y="% Cumulativa dos Pagamentos", data=topRevenueByCity)
ax.set_xlabel("Número de Cidades", fontsize = 14)
ax.set_ylabel("% Cumulativa dos Pagamentos", fontsize = 14)
ax.set_title("% Contribuição Cumulativa das Vendas nas Cidades" , fontsize = 18)

a = np.arange(1,len(topRevenueByCity)+1)
b = topRevenueByCity["% Cumulativa dos Pagamentos"]

plt.fill_between(a, b, 0,
                 where = (a >= 0) & (a <= 356),
                 color = 'r')

plt.fill_between(a, b, 0,
                 where = (a >= 0) & (a <= 43),
                 color = 'g')

ax.text(1000, 70, """43 Cidades (1.05% das cidades) contriburam 
para 50% das vendas.""", fontsize = 16)

ax.text(1000, 55, """356 Cidades (8.7% das cidades) contriburam 
para 80% das vendas.""", fontsize = 16)

Text(1000, 55, '356 Cidades (8.7% das cidades) contriburam \npara 80% das vendas.')


salesRegiao.head()


salesStatePercent = salesState.copy()


# Calculando porcentagem de cada estado em relação ao total de vendas
salesStatePercent['percent'] = (salesStatePercent['customer_id'] / salesStatePercent['customer_id'].sum()) * 100


salesStatePercent = salesStatePercent.sort_values(ascending = False, by = 'customer_id')


# Plot

# Definição dos dados no plot
plot_data = [go.Bar(x = salesStatePercent['customer_state'], 
                    y = salesStatePercent['percent'])]

# Layout
plot_layout = go.Layout(xaxis = {"type": "category"}, 
                        title = 'Total de Vendas por Estado em Porcentagem')

# Plot da figura
fig = go.Figure(data = plot_data, layout = plot_layout)
pyoff.iplot(fig)

# Upando no dash
if upload_to_dash:
    py.plot(fig, filename = 'total_vendas_estado_porcentagem', auto_open = False)


topRevenueByState = dtAll.groupby("customer_state")["payment_value"].sum().reset_index().sort_values("payment_value", ascending=False)
topRevenueByState["% of Total Payments"] = (topRevenueByState["payment_value"] / topRevenueByState["payment_value"].sum()) * 100
topRevenueByState["% Cumulativa dos Pagamentos"] = topRevenueByState["% of Total Payments"].cumsum()


topRevenueByState.head()


sns.set_style("ticks")

plt.figure(figsize=(12, 8))

ax = sns.lineplot(x=range(1,len(topRevenueByState)+1), y="% Cumulativa dos Pagamentos", data=topRevenueByState)
ax.set_xlabel("Número de Estados", fontsize = 14)
ax.set_ylabel("% Cumulativa dos Pagamentos", fontsize = 14)
ax.set_title("% Contribuição Cumulativa das Vendas nos Estados" , fontsize = 18)

a = np.arange(1,len(topRevenueByState)+1)
b = topRevenueByState["% Cumulativa dos Pagamentos"]

plt.fill_between(a, b, 0,
                 where = (a >= 0) & (a <= 7),
                 color = 'r')

plt.fill_between(a, b, 0,
                 where = (a >= 0) & (a <= 2),
                 color = 'g')

ax.text(12, 70, """2 Estados (7.41% dos Estados) contriburam 
para 51% das vendas.""", fontsize = 16)

ax.text(12, 55, """7 Estados (25.93% das Estados) contriburam 
para 81% das vendas.""", fontsize = 16)

Text(12, 55, '7 Estados (25.93% das Estados) contriburam \npara 81% das vendas.')


customerState = dtCustomers[['customer_unique_id', 'customer_state']].drop_duplicates().copy()


customerState = customerState.groupby('customer_state')['customer_unique_id'].count()


customerState = customerState.sort_values(ascending = False).reset_index()


customerState['percent'] = customerState['customer_unique_id'].apply(lambda x: str( round(x / sum(customerState['customer_unique_id']) * 100, 2) ) + '%')


customerState.head()


# Plot

# Definição dos dados no plot
plot_data = [go.Bar(x = customerState['customer_state'], 
                    y = customerState['customer_unique_id'],
                    text = customerState['percent'], textposition = 'outside')]

# Layout
plot_layout = go.Layout(xaxis = {"type": "category", 'title': 'Estados'},
                        yaxis = {'title': 'Vendas'},
                        title = 'Total de Clientes por Estado')

# Plot da figura
fig = go.Figure(data = plot_data, layout = plot_layout)
pyoff.iplot(fig)

# Upando no dash
if upload_to_dash:
    py.plot(fig, filename = 'total_clientes_estado', auto_open = False)


dtOrderItems.head()


dtProducts.head()


# Realizando o merge entre as duas tabelas
dtOrderProducts = pd.merge(dtOrderItems, dtProducts, on = 'product_id')


dtOrderProducts.head()


# Contando o numero de orders (não unicos) e relacionando a categoria
categoryProducts = dtOrderProducts.groupby('product_category_name')['order_id'].count().sort_values(ascending = False)


categoryProductsTop10 = categoryProducts.head(10).sort_values(ascending = False).reset_index()


categoryProducts = categoryProducts.reset_index()


categoryProductsTop10['percent'] = categoryProductsTop10['order_id'].apply(lambda x: str( round(x / sum(categoryProducts['order_id']) * 100, 2) ) + '%')


categoryProductsTop10 = categoryProductsTop10.sort_values(ascending = True, by = 'order_id')


# Plot

# Definição dos dados no plot
plot_data = [go.Bar(y = categoryProductsTop10['product_category_name'], 
                    x = categoryProductsTop10['order_id'], orientation = 'h',
                    text = categoryProductsTop10['percent'], textposition = 'outside')]

# Layout
plot_layout = go.Layout(yaxis = {"type": "category", 'title': 'Categorias'},
                        xaxis = {'title': 'Vendas'},
                        title = 'Top 10 Categorias Mais Vendidas')

# Plot da figura
fig = go.Figure(data = plot_data, layout = plot_layout)
pyoff.iplot(fig)

# Upando no dash
if upload_to_dash:
    py.plot(fig, filename = 'top_10_categorias_vendidas', auto_open = False)


topRevenueByCategory = dtAll.groupby("product_category_name")["payment_value"].sum().reset_index().sort_values("payment_value", ascending=False)
topRevenueByCategory["% of Total Payments"] = (topRevenueByCategory["payment_value"] / topRevenueByCategory["payment_value"].sum()) * 100
topRevenueByCategory["% Cumulativa dos Pagamentos"] = topRevenueByCategory["% of Total Payments"].cumsum()


sns.set_style("ticks")

plt.figure(figsize=(12, 8))

ax = sns.lineplot(x=range(1,len(topRevenueByCategory)+1), y="% Cumulativa dos Pagamentos", data=topRevenueByCategory)
ax.set_xlabel("Número de Categorias", fontsize = 14)
ax.set_ylabel("% Cumulativa dos Pagamentos", fontsize = 14)
ax.set_title("% Contribuição Cumulativa das Vendas nas Categorias" , fontsize = 18)

a = np.arange(1,len(topRevenueByCategory)+1)
b = topRevenueByCategory["% Cumulativa dos Pagamentos"]

plt.fill_between(a, b, 0,
                 where = (a >= 0) & (a <= 17),
                 color = 'r')

plt.fill_between(a, b, 0,
                 where = (a >= 0) & (a <= 7),
                 color = 'g')

ax.text(25, 70, """7 categorias (9.86% das categorias) contriburam 
para 51% das vendas.""", fontsize = 16)

ax.text(25, 55, """17 categorias (23.94% das categorias) contriburam 
para 81% das vendas.""", fontsize = 16)

Text(25, 55, '17 categorias (23.94% das categorias) contriburam \npara 81% das vendas.')


dtOrderItems.head()


dtOrderReviews.head()


reviewsOrders = pd.merge(dtOrderItems, dtOrderReviews, on = 'order_id')


reviewsOrders = reviewsOrders[['order_id', 'product_id', 'review_score']]


reviewsOrders = reviewsOrders.dropna()


reviewsOrders.head()


pedidosProduto = reviewsOrders.groupby('product_id')['order_id'].count().sort_values(ascending = False)


pedidosProduto.head()

product_id
aca2eb7d00ea1a7b8ebd4e68314663af    527
99a4788cb24856965c36a24e339b6058    491
422879e10f46682990de24d770e7f83d    487
389d119b48cf3043d311335e499d9c6b    392
368c6c730842d78016ad823897a372db    391
Name: order_id, dtype: int64


scoreProduto = reviewsOrders.groupby('product_id')['review_score'].mean().sort_values(ascending = False)


scoreProduto.head()

product_id
00066f42aeeb9f3007548bb9d3f33c38    5.0
86743ff92eee3d16b7df59cddd583b8c    5.0
868ceb027ab706a4dee42e2220006b85    5.0
868969d3a93aeeab7bfcd4fc3d3d65de    5.0
868766ed2172644fdd977d6bd395a107    5.0
Name: review_score, dtype: float64


scoreOrders = scoreProduto.copy()


scoreOrders = scoreOrders.to_frame()


scoreOrders = pd.merge(scoreProduto, pedidosProduto, on = 'product_id')


scoreOrders.head()


scoreOrders.columns = ['score_medio', 'numero_pedidos']


scoreOrders.head()


scoreOrders['score_medio'] = round(scoreOrders['score_medio'])


scoreOrders.sort_values(ascending = False, by = 'numero_pedidos').head()


meanScoreOrders = scoreOrders.groupby('score_medio')['numero_pedidos'].sum().reset_index()


meanScoreOrders['percent'] = meanScoreOrders['numero_pedidos'].apply(lambda x: str( round(x / sum(meanScoreOrders['numero_pedidos']) * 100, 2) ) + '%')


meanScoreOrders


# Plot

# Definição dos dados no plot
plot_data = [go.Bar(x = meanScoreOrders['score_medio'], 
                    y = meanScoreOrders['numero_pedidos'],
                    text = meanScoreOrders['percent'], textposition = 'outside')]

# Layout
plot_layout = go.Layout(xaxis = {"type": "category", 'title': 'Score'},
                        yaxis = {'title': 'Vendas'},
                        title = 'Vendas por Score (Score Médio por Produto)')

# Plot da figura
fig = go.Figure(data = plot_data, layout = plot_layout)
pyoff.iplot(fig)

# Upando no dash
if upload_to_dash:
    py.plot(fig, filename = 'vendas_por_score', auto_open = False)


dtOrderReviews.head()


dtOrderReviewsTemp = dtOrderReviews['review_comment_message']


dtOrderReviewsTemp = dtOrderReviewsTemp.dropna()


dtOrderReviewsTemp.head()

3                 Recebi bem antes do prazo estipulado.
4     Parabéns lojas lannister adorei comprar pela I...
9     aparelho eficiente. no site a marca do aparelh...
12      Mas um pouco ,travando...pelo valor ta Boa.\r\n
15    Vendedor confiável, produto ok e entrega antes...
Name: review_comment_message, dtype: object


def limpar_texto(text):
    # Convertendo para str
    text = str(text)
    
    # Remover caracteres non-ascii
    text = ''.join(caracter for caracter in text if ord(caracter) < 128)
    
    # Convertendo para lower case
    text = text.lower()
    
    # Removendo pontuação por expressão regular
    regex = re.compile('[' + re.escape(string.punctuation) + '\\r\\t\\n]')
    text = regex.sub(' ', str(text))
    
    # Carregando stopwords em português
    portuguese_stops = set(stopwords.words('portuguese'))
    
    # Removendo stopwords em português
    # Mantendo somente palavras que não são consideradas stopwords
    text = ' '.join(palavra for palavra in text.split() if palavra not in portuguese_stops)
    
    # Criando a estrutura baseada em uma wordnet para lemmatization
    wordnet_lemmatizer = WordNetLemmatizer()    
    # Aplicando Lemmatization
    text = ' '.join(wordnet_lemmatizer.lemmatize(palavra) for palavra in text.split())
    
    return text


dtOrderReviewsTemp = dtOrderReviewsTemp.to_frame()


dtOrderReviewsTemp


print(dtOrderReviewsTemp['review_comment_message'][9])

aparelho eficiente. no site a marca do aparelho esta impresso como 3desinfector e ao chegar esta com outro nome...atualizar com a marca correta uma vez que é o mesmo aparelho


dtOrderReviewsTratado = dtOrderReviewsTemp.copy()


%%time
print('Tratanto texto...')
dtOrderReviewsTratado['review_comment_message'] = dtOrderReviews['review_comment_message'].map(limpar_texto)
print('Texto tratado')

Tratanto texto...
Texto tratado
CPU times: total: 20.5 s
Wall time: 20.7 s


print(dtOrderReviewsTratado['review_comment_message'][9])

aparelho eficiente site marca aparelho impresso 3desinfector chegar outro nome atualizar marca correta vez aparelho


dtOrderReviewsTratado.head()


wordcloud = WordCloud(width = 600, height = 400).generate_from_text(\
                                                                ' '.join(dtOrderReviewsTratado['review_comment_message']))
plt.figure(figsize = (15, 10))
plt.imshow(wordcloud, interpolation = 'bilinear')
plt.axis('off')
plt.savefig('export/wordcloud.png')
plt.show()


dtGeolocation.head()


latIng = dtGeolocation[['geolocation_lat', 'geolocation_lng']]


latIng.head()


#geolocator = Nominatim(user_agent = 'Brazilian E-commerce Olist')


mapa = folium.Map([-10, -50], tiles = 'stamentoner', zoom_start = 5)


HeatMap(latIng, min_opacity = 0.2, radis = 2, blur = 10).add_to(mapa)

<folium.plugins.heat_map.HeatMap at 0x28792f9c250>


mapa.save('mapas/heatMap_ceps.html')


mapa = folium.Map([-10, -50], zoom_start = 5)


FastMarkerCluster(data = latIng).add_to(mapa)

<folium.plugins.fast_marker_cluster.FastMarkerCluster at 0x28786e30b80>


mapa.save('mapas/mapa_clusters_ceps.html')


# Juntando tabela de pedidos com itens
deliveryAnalysis = dtOrderItems.merge(dtOrders, how = 'left', on = 'order_id').merge(dtOrderReviews, how = 'left', on = 'order_id')


# Corrgindo tipos de colunas de data
deliveryAnalysis = deliveryAnalysis.assign(
    order_delivered_customer_date = pd.to_datetime(deliveryAnalysis["order_delivered_customer_date"]).dt.date,
    order_delivered_carrier_date = pd.to_datetime(deliveryAnalysis["order_delivered_carrier_date"]).dt.date,
    order_estimated_delivery_date = pd.to_datetime(deliveryAnalysis["order_estimated_delivery_date"]).dt.date,
    shipping_limit_date = pd.to_datetime(deliveryAnalysis["shipping_limit_date"]).dt.date,
    order_purchase_timestamp = pd.to_datetime(deliveryAnalysis['order_purchase_timestamp']).dt.date)


# Calculando tempo de entrega e estimado vs entregue
deliveryAnalysis = deliveryAnalysis.assign(delivered_days = (deliveryAnalysis['order_delivered_customer_date'] - deliveryAnalysis['order_purchase_timestamp']).dt.days,
                                 estimated_days_vs_delivery = (deliveryAnalysis['order_estimated_delivery_date'] - deliveryAnalysis['order_delivered_customer_date']).dt.days)


# Identificando status do pedido
deliveryAnalysis['delivery_status'] = np.where(deliveryAnalysis['estimated_days_vs_delivery'] < 0, 'Atrasado', 'No prazo')


conditions = [(deliveryAnalysis['estimated_days_vs_delivery'] < -30),
              (deliveryAnalysis['estimated_days_vs_delivery'] < -15),
              (deliveryAnalysis['estimated_days_vs_delivery'] < -10),
              (deliveryAnalysis['estimated_days_vs_delivery'] <= -5),
              (deliveryAnalysis['estimated_days_vs_delivery'] < 0)]
choices = ['Mais de 30 dias', 'Entre 15 e 30 dias', 'Entre 10 e 15 dias', 'Entre 5 e 10 dias','Menos de 5 dias']
deliveryAnalysis['delivery_status_detail'] = np.select(conditions, choices, default = 'No prazo')


deliveryAnalysis = deliveryAnalysis.drop_duplicates(keep = False, inplace = False)


deliveryAnalysisFiltered = deliveryAnalysis[['order_id','delivery_status']]
dlv = deliveryAnalysisFiltered.pivot_table(values = 'order_id',index = 'delivery_status', aggfunc = 'nunique')


# Filtrando status por atrasado para ver quanto tempo de atraso
late_deliver_df = deliveryAnalysis[deliveryAnalysis['delivery_status'] == 'Atrasado']
delayedDaysStatus = late_deliver_df.pivot_table(values = 'order_id', index = 'delivery_status_detail', aggfunc = 'nunique')
delayedDaysStatus = delayedDaysStatus.reindex(choices)


# Definição dos dados no plot
values = dlv.values

plot_data = [go.Pie(labels = dlv.index, values = dlv['order_id'], name = 'Status')]

# Layout
plot_layout = go.Layout(title={'text': '% de Pedidos Entregues no Período', 'y':0.9, 'x':0.5, 'xanchor': 'center', 'yanchor': 'top'})

# Plot da figura
fig = go.Figure(data = plot_data, layout = plot_layout)

fig.update_layout(legend_title_text='Status', legend=dict(
    yanchor="top",
    y = 0.80,
    xanchor="right",
    x = 0.7
))

pyoff.iplot(fig)

# Upando no dash
if upload_to_dash:
    py.plot(fig, filename = 'pedidos_atrasados_vs_no_prazo', auto_open = False)


# Definição dos dados no plot
plot_data = [go.Pie(labels = delayedDaysStatus.index, values = delayedDaysStatus['order_id'], name = 'NDiasAtraso', sort = False)]

# Layout
plot_layout = go.Layout(title={'text': '% Por Categoria de Dias de Atraso', 'y':0.9, 'x':0.5, 'xanchor': 'center', 'yanchor': 'top'})

# Plot da figura
fig = go.Figure(data = plot_data, layout = plot_layout)

fig.update_layout(legend_title_text='Dias atrasado', legend=dict(
    yanchor="top",
    y = 0.80,
    xanchor="right",
    x = 0.725
))


pyoff.iplot(fig)

# Upando no dash
if upload_to_dash:
    py.plot(fig, filename = 'categoria_dias_de_atraso', auto_open = False)


deliveryStatusByScore = pd.DataFrame()


for i in range(5):
    score = i + 1
    
    noPrazo = deliveryAnalysis.query("review_score == {0} and delivery_status == 'No prazo'".format(score)).count()[0]
    atrasado = deliveryAnalysis.query("review_score == {0} and delivery_status == 'Atrasado'".format(score)).count()[0]
    
    deliveryStatusByScore = deliveryStatusByScore.append(pd.DataFrame({'Score': [score], 'No prazo': [noPrazo], 'Atrasado': [atrasado]}))


# Definição dos dados no plot
plot_data = [go.Pie(labels = deliveryStatusByScore.Score, values = deliveryStatusByScore.Atrasado, name = 'NDiasAtraso', sort = False)]

# Layout
plot_layout = go.Layout(title={'text': '% de Pedidos Atrasados por Score', 'y':0.9, 'x':0.5, 'xanchor': 'center', 'yanchor': 'top'})

# Plot da figura
fig = go.Figure(data = plot_data, layout = plot_layout)

fig.update_layout(legend_title_text='Score', legend=dict(
    yanchor="top",
    y = 0.80,
    xanchor="right",
    x = 0.7
))

pyoff.iplot(fig)

# Upando no dash
if upload_to_dash:
    py.plot(fig, filename = 'pedidos_atrasados_por_score', auto_open = False)


deliveryAnalysisDelayed = deliveryAnalysis[deliveryAnalysis['estimated_days_vs_delivery'] < 0]


deliveryAnalysisDelayed = deliveryAnalysisDelayed[deliveryAnalysisDelayed['estimated_days_vs_delivery'] > -50]


fig = go.Figure()

fig.add_trace(go.Box(y= deliveryAnalysisDelayed[deliveryAnalysisDelayed['review_score'] == 1]['estimated_days_vs_delivery'], name = '1'))
fig.add_trace(go.Box(y= deliveryAnalysisDelayed[deliveryAnalysisDelayed['review_score'] == 2]['estimated_days_vs_delivery'], name = '2'))
fig.add_trace(go.Box(y= deliveryAnalysisDelayed[deliveryAnalysisDelayed['review_score'] == 3]['estimated_days_vs_delivery'], name = '3'))
fig.add_trace(go.Box(y= deliveryAnalysisDelayed[deliveryAnalysisDelayed['review_score'] == 4]['estimated_days_vs_delivery'], name = '4'))
fig.add_trace(go.Box(y= deliveryAnalysisDelayed[deliveryAnalysisDelayed['review_score'] == 5]['estimated_days_vs_delivery'], name = '5'))

fig.update_traces(boxpoints = 'outliers', jitter = 0)

fig.update_layout(legend_title_text = 'Score', title = 'Dias de Atraso por Score')

fig.show()


dtOrders.head()


reviewsOrders.head()


reviewOrdersTime = pd.merge(reviewsOrders, dtOrders, on = 'order_id')


reviewOrdersTime = reviewOrdersTime[reviewOrdersTime['order_status'] == 'delivered']


reviewOrdersTime = reviewOrdersTime[['order_id', 'product_id', 'review_score', 'order_delivered_carrier_date',\
                                     'order_delivered_customer_date', 'order_estimated_delivery_date']]


reviewOrdersTime.head()


# Mantendo somente Ano, Mes e Dia
reviewOrdersTime['order_delivered_carrier_date'] = \
    pd.to_datetime(reviewOrdersTime['order_delivered_carrier_date'], format = '%Y-%m-%d %H:%M:%S').dt.to_period('d')


# Mantendo somente Ano, Mes e Dia
reviewOrdersTime['order_delivered_customer_date'] = \
    pd.to_datetime(reviewOrdersTime['order_delivered_customer_date'], format = '%Y-%m-%d %H:%M:%S').dt.to_period('d')


# Mantendo somente Ano, Mes e Dia
reviewOrdersTime['order_estimated_delivery_date'] = \
    pd.to_datetime(reviewOrdersTime['order_estimated_delivery_date'], format = '%Y-%m-%d %H:%M:%S').dt.to_period('d')


reviewOrdersTime.head()


reviewOrdersTime.dtypes

order_id                            object
product_id                          object
review_score                         int64
order_delivered_carrier_date     period[D]
order_delivered_customer_date    period[D]
order_estimated_delivery_date    period[D]
dtype: object


print(reviewOrdersTime.isna().sum())

order_id                         0
product_id                       0
review_score                     0
order_delivered_carrier_date     2
order_delivered_customer_date    8
order_estimated_delivery_date    0
dtype: int64


reviewOrdersTime = reviewOrdersTime.dropna()


print(reviewOrdersTime.shape)

(110839, 6)


reviewOrdersTime['diff_delivery_and_estimate'] = reviewOrdersTime['order_estimated_delivery_date'] -\
                                           reviewOrdersTime['order_delivered_customer_date']


reviewOrdersTime['diff_delivery_and_estimate'] = \
                reviewOrdersTime['diff_delivery_and_estimate'].apply(lambda x: re.sub("[^0-9]", "", str(x)))


reviewOrdersTime['diff_delivery_and_estimate'] = pd.to_numeric(reviewOrdersTime['diff_delivery_and_estimate'])


reviewOrdersTime.head()


reviewOrdersTime.dtypes

order_id                            object
product_id                          object
review_score                         int64
order_delivered_carrier_date     period[D]
order_delivered_customer_date    period[D]
order_estimated_delivery_date    period[D]
diff_delivery_and_estimate         float64
dtype: object


reviewOrdersTime[reviewOrdersTime['diff_delivery_and_estimate'] < 0].head()


reviewOrdersTime[reviewOrdersTime['diff_delivery_and_estimate'] <= 0].head()


COLOR_5S = '#0571b0'
COLOR_1S = '#ca0020'
REVIEWS_PALETTE = sns.color_palette((COLOR_1S, '#d57b6f', '#c6c6c6', '#7f9abc', COLOR_5S))


resize_plot = lambda: plt.gcf().set_size_inches(12, 5)


reviewOrdersTime['order_delivered_customer_date'] = reviewOrdersTime['order_delivered_customer_date'].astype('str')


reviewOrdersTime.head()


fig = px.scatter(reviewOrdersTime, x="order_delivered_customer_date", y="diff_delivery_and_estimate", color = "review_score",
                 title="Diferença de Entrega entre Previsto e Real em Dias", labels = {'diff_delivery_and_estimate': 'Diferença de Dias entre Previsto e Real',
                                                                                                               'order_delivered_customer_date': 'Periodo', 'review_score': 'Score'})

fig.show()

# Upando no dash
#if upload_to_dash:
#    py.plot(fig, filename = 'diferenca_entrega_previst_real', auto_open = False)


sns.scatterplot(x = 'order_delivered_customer_date', y = 'diff_delivery_and_estimate',
               hue = 'review_score',
               palette = REVIEWS_PALETTE,
               data = reviewOrdersTime)
plt.xlabel('Delivered Date')
plt.ylabel('Difference Delivery and Estimate')
plt.title('Difference Delivery and Estimate (Negative values are delay)')
plt.xlim(reviewOrdersTime['order_delivered_customer_date'].min(), reviewOrdersTime['order_delivered_customer_date'].max())
plt.xticks([])
resize_plot()

C:\Users\herik\AppData\Local\Temp\ipykernel_2452\1907121668.py:9: MatplotlibDeprecationWarning:

Support for passing numbers through unit converters is deprecated since 3.5 and support will be removed two minor releases later; use Axis.convert_units instead.


customerNotUniqueState = dtCustomers[['customer_id', 'customer_state']].drop_duplicates().copy()


customerNotUniqueState = pd.merge(customerNotUniqueState, dtOrders, on = 'customer_id')


customerNotUniqueState = customerNotUniqueState[['customer_id', 'customer_state', 'order_id']]


customerNotUniqueState = pd.merge(customerNotUniqueState, dtOrderProducts , on = 'order_id')


customerNotUniqueStateFiltered = customerNotUniqueState[['customer_id', 'customer_state', 'order_id', 'product_category_name']]


customerNotUniqueStateFiltered.head()


customerNotUniqueStateFiltered[customerNotUniqueStateFiltered['customer_state'] == 'SP']['product_category_name'].\
    value_counts().head(5).reset_index()#.plot(kind = 'barh')


customerNotUniqueStateFiltered.head()


states = ['SP', 'RJ', 'MG', 'RS', 'PR', 'SC']
top_category = 5

col_nums = 2
row_nums = 3
row_ = 1
col_ = 1
maximumRange = 5500

fig = make_subplots(rows = row_nums, cols = col_nums, shared_yaxes = False, subplot_titles = states,
                   horizontal_spacing = 0.25)

for i, state in enumerate(states, 1):
    dtFiltrado = customerNotUniqueStateFiltered[customerNotUniqueStateFiltered['customer_state'] == states[i - 1]]['product_category_name'].\
    value_counts().head(top_category).sort_values(ascending = True).reset_index()
    
    fig.add_trace(go.Bar(x = dtFiltrado['product_category_name'],
                         y = dtFiltrado['index'], orientation = 'h'), row = row_, col = col_)
    fig.update_yaxes(showgrid = False, row = row_, col = col_, title = 'Categorias')
    fig.update_xaxes(showgrid = True, row = row_, col = col_, title = 'Vendas')
    fig.update_xaxes(range = [0, maximumRange])
    
    col_ += 1
    if col_ > col_nums:
        col_ = 1
        row_ += 1


fig.update_layout(showlegend = False, title_text = "Categorias Mais Vendidas por Estado", height = 700)
fig.show()

# Upando no dash
if upload_to_dash:
    py.plot(fig, filename = 'categorias_mais_vendidas_por_estado', auto_open = False)


states = ['SP', 'RJ', 'MG', 'RS', 'PR', 'SC']
top_category = 5

col_nums = 2
row_nums = 3
row_ = 1
col_ = 1
maximumRange = 5500

fig = make_subplots(rows = row_nums, cols = col_nums, shared_yaxes = False, subplot_titles = states,
                   horizontal_spacing = 0.25)

topGeneralProducts = customerNotUniqueStateFiltered['product_category_name'].value_counts().head(5).reset_index()['index'].tolist()

for i, state in enumerate(states, 1):
    dtFiltrado = customerNotUniqueStateFiltered[customerNotUniqueStateFiltered['customer_state'] == states[i - 1]]
    dtFiltrado = dtFiltrado[dtFiltrado['product_category_name'].isin(topGeneralProducts)]['product_category_name'].value_counts().head(5).reset_index().sort_values('index')
    
    fig.add_trace(go.Bar(x = dtFiltrado['product_category_name'],
                         y = dtFiltrado['index'], orientation = 'h'), row = row_, col = col_)
    fig.update_yaxes(showgrid = False, row = row_, col = col_, title = 'Categorias')
    fig.update_xaxes(showgrid = True, row = row_, col = col_, title = 'Vendas')
    fig.update_xaxes(range = [0, maximumRange])
    
    col_ += 1
    if col_ > col_nums:
        col_ = 1
        row_ += 1


fig.update_layout(showlegend = False, title_text = "Categorias Mais Vendidas nos Estados de Maior Demanda", height = 700)
fig.show()

# Upando no dash
if upload_to_dash:
    py.plot(fig, filename = 'categorias_mais_vendidas_nos_estados_de_maior_demanda', auto_open = False)


dtFiltrado = pd.merge(dtOrdersAjustado, customerNotUniqueStateFiltered, on = 'order_id')


dtFiltrado = dtFiltrado[['order_purchase_timestamp', 'product_category_name']]


dtFiltrado2 = dtFiltrado.copy()


dtFiltrado2['order_purchase_timestamp'] = dtFiltrado2['order_purchase_timestamp'].dt.date


dtFiltrado2 = dtFiltrado2.sort_values('order_purchase_timestamp')


dtFiltrado2.head()


dtFiltrado2 = dtFiltrado2.groupby(['order_purchase_timestamp','product_category_name']).agg(pedidos=('product_category_name', 'count'))


dtFiltrado2 = dtFiltrado2.reset_index()


dtFiltrado2['order_purchase_timestamp'] = pd.to_datetime(dtFiltrado2['order_purchase_timestamp'], format = '%Y-%m-%d')


# Remove outliers
dtFiltrado2 = dtFiltrado2[(dtFiltrado2['order_purchase_timestamp'] > "2017-01-01") & (dtFiltrado2['order_purchase_timestamp'] < "2018-08-07")]
dtFiltrado2 = dtFiltrado2[(dtFiltrado2['order_purchase_timestamp'] != "2017-11-24")]


dtFiltrado2.head()


topGeneralProducts = customerNotUniqueStateFiltered['product_category_name'].value_counts().head(5).reset_index()['index'].tolist()


fig = go.Figure()

for i in topGeneralProducts:
    fig.add_trace(go.Scatter(x=dtFiltrado2[dtFiltrado2['product_category_name']== i]['order_purchase_timestamp'], y=dtFiltrado2[dtFiltrado2['product_category_name']== i]['pedidos'],\
                    mode = 'lines',\
                    name = i, connectgaps=True,\
                    line=dict(width=1.5, dash='dash')))
    
fig.update_layout(title_text = "Categorias Mais Vendidas entre 2016 e 2018")
fig.update_yaxes(title = 'Vendas')
fig.update_xaxes(title = 'Período')

fig.show()


dtCustomers.head()


dtOrders.head()


dtCustomersOrders = pd.merge(dtCustomers, dtOrders, on = 'customer_id')


# Mantendo somente Ano e Mes
dtCustomersOrders['order_purchase_timestamp'] =\
        pd.to_datetime(dtCustomersOrders['order_purchase_timestamp'], format = '%Y-%m-%d %H:%M:%S').dt.to_period('M')


dtCustomersOrders = dtCustomersOrders[['order_purchase_timestamp', 'customer_id']]


dtCustomersOrders.columns = ['AnoMes', 'customer_id']


dtCustomersOrders.head()


dtActiveCustomers = dtCustomersOrders.groupby('AnoMes')['customer_id'].nunique().reset_index()


dtActiveCustomers['AnoMes'] = dtActiveCustomers['AnoMes'].astype(str)


dtActiveCustomers['percent'] = dtActiveCustomers['customer_id'].apply(lambda x: str( round(x / sum(dtActiveCustomers['customer_id']) * 100, 2) ) + '%')


dtActiveCustomers.head()


# Plot

# Definição dos dados no plot
plot_data = [go.Bar(x = dtActiveCustomers['AnoMes'], 
                    y = dtActiveCustomers['customer_id'],
                    text = dtActiveCustomers['customer_id'], textposition = 'outside', textfont=dict(size = 14
            ))]

# Layout
plot_layout = go.Layout(xaxis = {"type": "category", 'title': 'Periodo'},
                        yaxis = {'title': 'Vendas'},
                        title = 'Clientes Ativos Por Mês')

# Plot da figura
fig = go.Figure(data = plot_data, layout = plot_layout)

pyoff.iplot(fig)

# Upando no dash
if upload_to_dash:
    py.plot(fig, filename = 'clientes_ativos_por_mes', auto_open = False)


dtCustomersOrders = pd.merge(dtCustomers, dtOrders, on = 'customer_id')


# Mantendo somente Ano e Mes
dtCustomersOrders['order_purchase_timestamp'] =\
        pd.to_datetime(dtCustomersOrders['order_purchase_timestamp'], format = '%Y-%m-%d %H:%M:%S')


dtCustomersOrders['order_purchase_timestamp'] =\
            dtCustomersOrders['order_purchase_timestamp'].map(lambda date: 100 * date.year + date.month)


dtCustomersOrders.head()


dtOrderPrice = dtOrderItems.groupby('order_id').agg({'price': sum}).reset_index()


dtOrderPrice.head()


dtCustomersOrders = pd.merge(dtCustomersOrders, dtOrderPrice, on = 'order_id')


dtCustomersOrders.head()


dtPurchaseCustomers = dtCustomersOrders.groupby(['customer_unique_id', 'order_purchase_timestamp'])['price'].sum().reset_index()


dtPurchaseCustomers.head()


df_ret = pd.crosstab(dtPurchaseCustomers['customer_unique_id'], dtPurchaseCustomers['order_purchase_timestamp']).reset_index()


df_ret.head()


# Extraímos os meses
meses = df_ret.columns[2:]


meses

Index([201610, 201612, 201701, 201702, 201703, 201704, 201705, 201706, 201707,
       201708, 201709, 201710, 201711, 201712, 201801, 201802, 201803, 201804,
       201805, 201806, 201807, 201808, 201809],
      dtype='object', name='order_purchase_timestamp')


'''# O loop abaixo vai calcular a retenção ao longo dos meses

# Lista para gravar o resulatdo
lista_ret = []

# Loop
for i in range(len(meses)-1):
    dados_retencao = {}
    mes_corrente = meses[i+1]
    mes_anterior = meses[i]
    dados_retencao['AnoMes'] = int(mes_corrente)
    dados_retencao['TotalUser'] = df_ret[mes_corrente].sum()
    dados_retencao['TotalRetido'] = df_ret[(df_ret[mes_corrente] > 0) & (df_ret[mes_anterior] > 0)][mes_corrente].sum()
    lista_ret.append(dados_retencao)'''

"# O loop abaixo vai calcular a retenção ao longo dos meses\n\n# Lista para gravar o resulatdo\nlista_ret = []\n\n# Loop\nfor i in range(len(meses)-1):\n    dados_retencao = {}\n    mes_corrente = meses[i+1]\n    mes_anterior = meses[i]\n    dados_retencao['AnoMes'] = int(mes_corrente)\n    dados_retencao['TotalUser'] = df_ret[mes_corrente].sum()\n    dados_retencao['TotalRetido'] = df_ret[(df_ret[mes_corrente] > 0) & (df_ret[mes_anterior] > 0)][mes_corrente].sum()\n    lista_ret.append(dados_retencao)"


# O loop abaixo vai calcular a retenção ao longo dos meses

# Lista para gravar o resulatdo
lista_ret = []

# Loop
for i in range(len(meses)-3):
    dados_retencao = {}
    mes_corrente = meses[i + 3]
    mes_anterior0 = meses[i]
    mes_anterior1 = meses[i + 1]
    mes_anterior2 = meses[i + 2]
    dados_retencao['AnoMes'] = int(mes_corrente)
    dados_retencao['TotalUser'] = df_ret[mes_corrente].sum()
    dados_retencao['TotalRetido'] = df_ret[(df_ret[mes_corrente] > 0) & ((df_ret[mes_anterior0] > 0) | (df_ret[mes_anterior1] > 0) | (df_ret[mes_anterior2] > 0))][mes_corrente].sum()
    lista_ret.append(dados_retencao)


df_ret_final = pd.DataFrame(lista_ret)


df_ret_final.head()


# Calculo do indicador
df_ret_final['TaxaRetencao'] = df_ret_final['TotalRetido'] / df_ret_final['TotalUser']


df_ret_final.head()


# Plot

# Definição dos dados no plot
plot_data = [go.Scatter(x = df_ret_final.query('AnoMes < 201809')['AnoMes'],
                        y = df_ret_final.query('AnoMes < 201809')['TaxaRetencao'],
                        name = "taxa")]

# Layout
plot_layout = go.Layout(xaxis = {"type": "category", 'title': 'Periodo'},
                        yaxis = {'title': 'Taxa de Retenção'},
                        title = 'Taxa Mensal de Retenção de Clientes')

# Plot da figura
fig = go.Figure(data = plot_data, layout = plot_layout)

fig.add_annotation(
        x = 16,
        y = 0.014,
        xref="x",
        yref="y",
        text="Maior Taxa de Retenção",
        showarrow = True,
        font=dict(
            family="Courier New, monospace",
            size=16,
            color="#000"
            ),
        align="center",
        arrowhead=2,
        arrowsize=1,
        arrowwidth=2,
        arrowcolor="#636363",
        ax=20,
        ay=-30,
        bordercolor="#636363",
        borderwidth=1.5,
        borderpad=4,
        bgcolor="#fff",
        opacity=1
        )

pyoff.iplot(fig)

# Upando no dash
if upload_to_dash:
    py.plot(fig, filename = 'taxa_mensal_retencao_clientes', auto_open = False)


dtCustomersOrders = pd.merge(dtCustomers, dtOrders, on = 'customer_id')


dtCustomersOrders.head()


# Mantendo somente Ano, Mes e Dia
dtCustomersOrders['order_purchase_timestamp'] = pd.to_datetime(dtCustomersOrders['order_purchase_timestamp'],\
                                                        format = '%Y-%m-%d %H:%M:%S')


# Vamos encontrar a data de menor volume de compras de cada cliente
dtMinPurchase = dtCustomersOrders.groupby('customer_unique_id')["order_purchase_timestamp"].min().reset_index()


# Vamos extrair o mês em que ocorreu o menor volume de compras de cada cliente
dtMinPurchase['Mes_Menor_Compra_Mensal'] =\
dtMinPurchase['order_purchase_timestamp'].map(lambda date: 100 * date.year + date.month)


dtMinPurchase.head()


dtCustomersOrders['order_purchase_timestamp'] =\
dtCustomersOrders['order_purchase_timestamp'].map(lambda date: 100 * date.year + date.month)


dtCustomersOrders.head()


dtPurchases = pd.merge(dtCustomersOrders, dtMinPurchase, on = "customer_unique_id")


dtPurchases.head()


dtPurchases['TipoUsuario'] = 'Novo'


dtPurchases['TipoUsuario'].value_counts()

Novo    99441
Name: TipoUsuario, dtype: int64


dtPurchases.loc[dtPurchases['order_purchase_timestamp_x'] > dtPurchases['Mes_Menor_Compra_Mensal'], 'TipoUsuario'] = 'Antigo'


dtPurchases['TipoUsuario'].value_counts()

Novo      97418
Antigo     2023
Name: TipoUsuario, dtype: int64


dtPurchases.head()


dtOrderPrice.head()


dtPurchases = pd.merge(dtPurchases, dtOrderPrice, on = 'order_id')


dtRevenuesCustomers = dtPurchases.groupby(['order_purchase_timestamp_x', 'TipoUsuario'])['price'].sum().reset_index()


dtRevenuesCustomers.columns = ['AnoMes', 'TipoUsuario', 'Faturamento']


dtRevenuesCustomers = dtRevenuesCustomers.append({'AnoMes': '201609',\
                                                  'TipoUsuario': 'Antigo', 'Faturamento': 0}, ignore_index = True)
dtRevenuesCustomers = dtRevenuesCustomers.append({'AnoMes': '201610',\
                                                  'TipoUsuario': 'Antigo', 'Faturamento': 0}, ignore_index = True)
dtRevenuesCustomers = dtRevenuesCustomers.append({'AnoMes': '201612',\
                                                  'TipoUsuario': 'Antigo', 'Faturamento': 0}, ignore_index = True)
dtRevenuesCustomers = dtRevenuesCustomers.append({'AnoMes': '201809',\
                                                  'TipoUsuario': 'Novo', 'Faturamento': 0}, ignore_index = True)


dtRevenuesCustomers['AnoMes'] = dtRevenuesCustomers['AnoMes'].astype(int)


dtRevenuesCustomers = dtRevenuesCustomers.sort_values(ascending = True, by = 'AnoMes')


dtRevenuesCustomers.head()


# Plot

# Definição dos dados no plot
plot_data = [go.Scatter(x = dtRevenuesCustomers.query("TipoUsuario == 'Antigo'")['AnoMes'],
                        y = dtRevenuesCustomers.query("TipoUsuario == 'Antigo'")['Faturamento'],
                        name = 'Cliente Antigo'),
             go.Scatter(x = dtRevenuesCustomers.query("TipoUsuario == 'Novo'")['AnoMes'],
                        y = dtRevenuesCustomers.query("TipoUsuario == 'Novo'")['Faturamento'],
                        name = 'Cliente Novo')]

# Layout
plot_layout = go.Layout(xaxis = {"type": "category", 'title': 'Periodo'},
                        yaxis = {'title': 'Faturamento'}, 
                        title = 'Diferença de Faturamento ao Longo do Tempo Entre Clientes Novos e Antigos')

# Plot da figura
fig = go.Figure(data = plot_data, layout = plot_layout)
pyoff.iplot(fig)

# Upando no dash
if upload_to_dash:
    py.plot(fig, filename = 'diferenca_faturamento_clientes_novos_e_antigos', auto_open = False)


agg_name = 'geolocation_zip_code_prefix'


def plot_map(data, label, agg_data, agg_name, cmap):
    url="http://server.arcgisonline.com/ArcGIS/rest/services/Canvas/World_Dark_Gray_Base/MapServer/tile/{Z}/{Y}/{X}.png"
    
    geomap = gv.WMTS(url)
    points = hv.Points(gv.Dataset(data, kdims=['x', 'y'], vdims=[agg_name]))
    agg = datashade(points, element_type=gv.Image, aggregator=agg_data, cmap=cmap)
    zip_codes = dynspread(agg, threshold=T, max_px=PX)
    hover = hv.util.Dynamic(rasterize(points, aggregator=agg_data, width=50, height=25, streams=[RangeXY]), operation=hv.QuadMesh)
    hover = hover.options(cmap=cmap)
    img = geomap * zip_codes * hover
    img = img.relabel(label)
    
    return img


background = "black"
cm = partial(colormap_select, reverse=(background!="black"))
export = partial(export_image, background = background, export_path="export")
W = 700 

def create_map(data, cmap, data_agg, export_name='img'):
    pad = (data.x.max() - data.x.min())/50
    x_range, y_range = ((data.x.min() - pad, data.x.max() + pad), 
                             (data.y.min() - pad, data.y.max() + pad))

    ratio = (y_range[1] - y_range[0]) / (x_range[1] - x_range[0])

    plot_width  = int(W)
    plot_height = int(plot_width * ratio)
    if ratio > 1.5:
        plot_height = 550
        plot_width = int(plot_height / ratio)
        
    cvs = ds.Canvas(plot_width=plot_width, plot_height=plot_height, x_range=x_range, y_range=y_range)

    agg = cvs.points(data, 'x', 'y', data_agg)
    img = tf.shade(agg, cmap=cmap, how='eq_hist')
    return export(img, export_name)


dtGeolocation.head()


dtGeolocation.dtypes

geolocation_zip_code_prefix     object
geolocation_lat                float64
geolocation_lng                float64
geolocation_city                object
geolocation_state               object
regiao                          object
dtype: object


dtGeoMod = dtGeolocation.copy()


dtGeoMod['geolocation_zip_code_prefix_1_digits'] = dtGeoMod['geolocation_zip_code_prefix'].str[0:1]
dtGeoMod['geolocation_zip_code_prefix_2_digits'] = dtGeoMod['geolocation_zip_code_prefix'].str[0:2]
dtGeoMod['geolocation_zip_code_prefix_3_digits'] = dtGeoMod['geolocation_zip_code_prefix'].str[0:3]
dtGeoMod['geolocation_zip_code_prefix_4_digits'] = dtGeoMod['geolocation_zip_code_prefix'].str[0:4]


dtGeoMod.head()


# Removendo outliers, iremos utilizar somente coordenadas no Brasil

# Pontos mais extremos do Brasil
dtGeoMod = dtGeoMod[dtGeoMod['geolocation_lat'] <= 5.27438888]
dtGeoMod = dtGeoMod[dtGeoMod['geolocation_lng'] >= -73.98283055]
dtGeoMod = dtGeoMod[dtGeoMod['geolocation_lat'] >= -33.75116944]
dtGeoMod = dtGeoMod[dtGeoMod['geolocation_lng'] <=  -34.79314722]


x, y = webm(dtGeoMod['geolocation_lng'], dtGeoMod['geolocation_lat'])
dtGeoMod['x'] = pd.Series(x)
dtGeoMod['y'] = pd.Series(y)


dtGeoMod.head()


dtGeoMod['geolocation_zip_code_prefix'] = dtGeoMod['geolocation_zip_code_prefix'].astype(int)
dtGeoMod['geolocation_zip_code_prefix_1_digits'] = dtGeoMod['geolocation_zip_code_prefix_1_digits'].astype(int)
dtGeoMod['geolocation_zip_code_prefix_2_digits'] = dtGeoMod['geolocation_zip_code_prefix_2_digits'].astype(int)
dtGeoMod['geolocation_zip_code_prefix_3_digits'] = dtGeoMod['geolocation_zip_code_prefix_3_digits'].astype(int)
dtGeoMod['geolocation_zip_code_prefix_4_digits'] = dtGeoMod['geolocation_zip_code_prefix_4_digits'].astype(int)


plot_map(dtGeoMod, 'CEP no Brasil', ds.min(agg_name), agg_name, cmap=rainbow)


create_map(dtGeoMod, rainbow, ds.mean(agg_name),'CEP_brasil')


plot_map(dtGeoMod[dtGeoMod['geolocation_state'] == 'SP'], 'CEP no Estado de São Paulo', ds.min(agg_name), agg_name, cmap = rainbow)


create_map(dtGeoMod[dtGeoMod['geolocation_state'] == 'SP'], rainbow, ds.mean(agg_name), 'ceps_SP_estado')


plot_map(dtGeoMod[dtGeoMod['geolocation_city'] == 'sao paulo'], 'CEP na Cidade de São Paulo', ds.min(agg_name), agg_name, cmap = rainbow)


create_map(dtGeoMod[dtGeoMod['geolocation_city'] == 'sao paulo'], rainbow, ds.mean(agg_name), 'ceps_SP_cidade')


plot_map(dtGeoMod[dtGeoMod['geolocation_state'] == 'SC'], 'CEP no Estado de Santa Catarina', ds.min(agg_name), agg_name, cmap = rainbow)


create_map(dtGeoMod[dtGeoMod['geolocation_state'] == 'SC'], rainbow, ds.mean(agg_name), 'ceps_SC_estado')


plot_map(dtGeoMod[dtGeoMod['geolocation_city'] == 'itajai'], 'CEP na Cidade de Itajaí', ds.min(agg_name), agg_name, cmap = rainbow)


create_map(dtGeoMod[dtGeoMod['geolocation_city'] == 'itajai'], rainbow, ds.mean(agg_name), 'ceps_itajai_cidade')


plot_map(dtGeoMod[dtGeoMod['geolocation_city'] == 'rio de janeiro'], 'CEP na Cidade do Rio De Janeiro', ds.min(agg_name), agg_name, cmap = rainbow)


create_map(dtGeoMod[dtGeoMod['geolocation_city'] == 'rio de janeiro'], rainbow, ds.mean(agg_name), 'ceps_rio_de_janeiro_cidade')


dtCustomers.head()


dtCustomersTemp = dtCustomers.copy()


dtCustomersTemp['customer_zip_code_prefix_3_digits'] = dtCustomersTemp['customer_zip_code_prefix'].str[0:3]
dtCustomersTemp['customer_zip_code_prefix_3_digits'] = dtCustomersTemp['customer_zip_code_prefix_3_digits'].astype(int)

dtGeoModBR = dtGeoMod.set_index('geolocation_zip_code_prefix_3_digits').copy()


dtOrderItems.head()


dtOrdersTemp = dtOrders.merge(dtOrderItems, on='order_id')
dtOrdersTemp = dtOrdersTemp.merge(dtCustomersTemp, on='customer_id')
dtOrdersTemp = dtOrdersTemp.merge(dtOrderReviews, on = 'order_id')


dtFiltrado = dtOrdersTemp.groupby('customer_zip_code_prefix_3_digits')['price'].sum().to_frame()
revenue = dtGeoModBR.join(dtFiltrado)
agg_name = 'revenue'
revenue[agg_name] = revenue.price / 1000


plot_map(revenue, 'Faturamento no Brasil (Em milhares de R$)', ds.mean(agg_name), agg_name, cmap=fire)


create_map(revenue, fire, ds.mean(agg_name), 'faturamento_brasil')


dtFiltrado = dtOrdersTemp.groupby('order_id').agg({'price': 'sum', 'customer_zip_code_prefix_3_digits': 'max'})
dtFiltrado = dtFiltrado.groupby('customer_zip_code_prefix_3_digits')['price'].mean().to_frame()
avg_ticket = dtGeoModBR.join(dtFiltrado)
agg_name = 'avg_ticket'
avg_ticket[agg_name] = avg_ticket.price


plot_map(avg_ticket, 'Ticket Medio em R$ no Brasil', ds.mean(agg_name), agg_name, cmap=bgy)


create_map(avg_ticket, bgy, ds.mean('avg_ticket'), 'ticket_medio_brasil')


dtFiltrado = dtOrdersTemp.groupby('order_id').agg({'price': 'sum', 'freight_value': 'sum', 'customer_zip_code_prefix_3_digits': 'max'})
agg_name = 'freight_ratio'
dtFiltrado[agg_name] = dtFiltrado['freight_value'] / dtFiltrado['price']
dtFiltrado = dtFiltrado.groupby('customer_zip_code_prefix_3_digits')[agg_name].mean().to_frame()
freight_ratio = dtGeoModBR.join(dtFiltrado)


plot_map(freight_ratio, 'Taxa de Transporte em Relação ao Pedido no Brasil', ds.mean(agg_name), agg_name, cmap=bgy)


create_map(freight_ratio, bgy, ds.mean('freight_ratio'), 'taxa_transporte_em_relacao_ao_pedido_brasil')


dtOrdersTemp['order_delivered_customer_date'] = pd.to_datetime(dtOrdersTemp['order_delivered_customer_date'])
dtOrdersTemp['order_estimated_delivery_date'] = pd.to_datetime(dtOrdersTemp['order_estimated_delivery_date'])
dtOrdersTemp['order_delivered_carrier_date'] = pd.to_datetime(dtOrdersTemp['order_delivered_carrier_date'])
dtOrdersTemp['actual_delivery_time'] = dtOrdersTemp['order_delivered_customer_date'] - dtOrdersTemp['order_delivered_carrier_date']
dtOrdersTemp['actual_delivery_time'] = dtOrdersTemp['actual_delivery_time'].dt.days


dtFiltrado = dtOrdersTemp.groupby('customer_zip_code_prefix_3_digits')['actual_delivery_time'].mean().to_frame()
delivery_time = dtGeoModBR.join(dtFiltrado)
agg_name = 'avg_delivery_time'
delivery_time[agg_name] = delivery_time['actual_delivery_time']


plot_map(delivery_time, 'Tempo Medio de Entrega em Dias no Brasil', ds.mean(agg_name), agg_name, cmap=bjy)


create_map(delivery_time, bjy, ds.mean(agg_name), 'tempo_medio_entrega_dias_brasil')


dtFiltrado = dtOrdersTemp.groupby('customer_zip_code_prefix_3_digits')['review_score'].mean().to_frame()
score = dtGeoModBR.join(dtFiltrado)
agg_name = 'avg_score'
score[agg_name] = score['review_score']


plot_map(score, 'Review Médio No Brasil', ds.mean(agg_name), agg_name, cmap=bgy)


create_map(score, bgy, ds.mean(agg_name), 'review_medio_brasil')

	customer_id	customer_unique_id	customer_zip_code_prefix	customer_city	customer_state	regiao
0	06b8999e2fba1a1fbc88172c00ba8bc7	861eff4711a542e4b93843c6dd7febb0	14409	franca	SP	Sudeste
1	18955e83d337fd6b2def6b18a428ac77	290c77bc529b7ac935b93aa66c333dc3	09790	sao bernardo do campo	SP	Sudeste
2	4e7b3e00288586ebd08712fdd0374a03	060e732b5b29e8181a18229c7b0b2b5e	01151	sao paulo	SP	Sudeste
3	b2b6027bc5c5109e529d4dc6358b12c3	259dac757896d24d7702b9acbbff3f3c	08775	mogi das cruzes	SP	Sudeste
4	4f2d8ab171c80ec8364f7c12e35b23ad	345ecd01c38d18a9036ed96c73b8d066	13056	campinas	SP	Sudeste

	customer_id	customer_unique_id	customer_zip_code_prefix	customer_city	customer_state	regiao
0	06b8999e2fba1a1fbc88172c00ba8bc7	861eff4711a542e4b93843c6dd7febb0	14409	franca	SP	Sudeste
1	18955e83d337fd6b2def6b18a428ac77	290c77bc529b7ac935b93aa66c333dc3	09790	sao bernardo do campo	SP	Sudeste
2	4e7b3e00288586ebd08712fdd0374a03	060e732b5b29e8181a18229c7b0b2b5e	01151	sao paulo	SP	Sudeste
3	b2b6027bc5c5109e529d4dc6358b12c3	259dac757896d24d7702b9acbbff3f3c	08775	mogi das cruzes	SP	Sudeste
4	4f2d8ab171c80ec8364f7c12e35b23ad	345ecd01c38d18a9036ed96c73b8d066	13056	campinas	SP	Sudeste

	order_id	customer_id	order_status	order_purchase_timestamp	order_approved_at	order_delivered_carrier_date	order_delivered_customer_date	order_estimated_delivery_date
0	e481f51cbdc54678b7cc49136f2d6af7	9ef432eb6251297304e76186b10a928d	delivered	2017-10-02 10:56:33	2017-10-02 11:07:15	2017-10-04 19:55:00	2017-10-10 21:25:13	2017-10-18 00:00:00
1	53cdb2fc8bc7dce0b6741e2150273451	b0830fb4747a6c6d20dea0b8c802d7ef	delivered	2018-07-24 20:41:37	2018-07-26 03:24:27	2018-07-26 14:31:00	2018-08-07 15:27:45	2018-08-13 00:00:00
2	47770eb9100c2d0c44946d9cf07ec65d	41ce2a54c0b03bf3443c3d931a367089	delivered	2018-08-08 08:38:49	2018-08-08 08:55:23	2018-08-08 13:50:00	2018-08-17 18:06:29	2018-09-04 00:00:00
3	949d5b44dbf5de918fe9c16f97b45f8a	f88197465ea7920adcdbec7375364d82	delivered	2017-11-18 19:28:06	2017-11-18 19:45:59	2017-11-22 13:39:59	2017-12-02 00:28:42	2017-12-15 00:00:00
4	ad21c59c0840e6cb83a9ceb5573f8159	8ab97904e6daea8866dbdbc4fb7aad2c	delivered	2018-02-13 21:18:39	2018-02-13 22:20:29	2018-02-14 19:46:34	2018-02-16 18:17:02	2018-02-26 00:00:00

	order_id	customer_id	order_status	order_purchase_timestamp	order_approved_at	order_delivered_carrier_date	order_delivered_customer_date	order_estimated_delivery_date
0	e481f51cbdc54678b7cc49136f2d6af7	9ef432eb6251297304e76186b10a928d	delivered	2017-10-02 10:56:33	2017-10-02 11:07:15	2017-10-04 19:55:00	2017-10-10 21:25:13	2017-10-18
1	53cdb2fc8bc7dce0b6741e2150273451	b0830fb4747a6c6d20dea0b8c802d7ef	delivered	2018-07-24 20:41:37	2018-07-26 03:24:27	2018-07-26 14:31:00	2018-08-07 15:27:45	2018-08-13
2	47770eb9100c2d0c44946d9cf07ec65d	41ce2a54c0b03bf3443c3d931a367089	delivered	2018-08-08 08:38:49	2018-08-08 08:55:23	2018-08-08 13:50:00	2018-08-17 18:06:29	2018-09-04
3	949d5b44dbf5de918fe9c16f97b45f8a	f88197465ea7920adcdbec7375364d82	delivered	2017-11-18 19:28:06	2017-11-18 19:45:59	2017-11-22 13:39:59	2017-12-02 00:28:42	2017-12-15
4	ad21c59c0840e6cb83a9ceb5573f8159	8ab97904e6daea8866dbdbc4fb7aad2c	delivered	2018-02-13 21:18:39	2018-02-13 22:20:29	2018-02-14 19:46:34	2018-02-16 18:17:02	2018-02-26

	order_id	customer_id	order_status	order_purchase_timestamp	order_approved_at	order_delivered_carrier_date	order_delivered_customer_date	order_estimated_delivery_date
6	136cce7faa42fdb2cefd53fdc79a6098	ed0271e0b7da060a393796590e7b737a	invoiced	2017-04-11 12:22:08	2017-04-13 13:25:17	NaT	NaT	2017-05-09
44	ee64d42b8cf066f35eac1cf57de1aa85	caded193e8e47b8362864762a83db3c5	shipped	2018-06-04 16:44:48	2018-06-05 04:31:18	2018-06-05 14:32:00	NaT	2018-06-28
103	0760a852e4e9d89eb77bf631eaaf1c84	d2a79636084590b7465af8ab374a8cf5	invoiced	2018-08-03 17:44:42	2018-08-07 06:15:14	NaT	NaT	2018-08-21
128	15bed8e2fec7fdbadb186b57c46c92f2	f3f0e613e0bdb9c7cee75504f0f90679	processing	2017-09-03 14:22:03	2017-09-03 14:30:09	NaT	NaT	2017-10-03
154	6942b8da583c2f9957e990d028607019	52006a9383bf149a4fb24226b173106f	shipped	2018-01-10 11:33:07	2018-01-11 02:32:30	2018-01-11 19:39:23	NaT	2018-02-07
...	...	...	...	...	...	...	...	...
99283	3a3cddda5a7c27851bd96c3313412840	0b0d6095c5555fe083844281f6b093bb	canceled	2018-08-31 16:13:44	NaT	NaT	NaT	2018-10-01
99313	e9e64a17afa9653aacf2616d94c005b8	b4cd0522e632e481f8eaf766a2646e86	processing	2018-01-05 23:07:24	2018-01-09 07:18:05	NaT	NaT	2018-02-06
99347	a89abace0dcc01eeb267a9660b5ac126	2f0524a7b1b3845a1a57fcf3910c4333	canceled	2018-09-06 18:45:47	NaT	NaT	NaT	2018-09-27
99348	a69ba794cc7deb415c3e15a0a3877e69	726f0894b5becdf952ea537d5266e543	unavailable	2017-08-23 16:28:04	2017-08-28 15:44:47	NaT	NaT	2017-09-15
99415	5fabc81b6322c8443648e1b21a6fef21	32c9df889d41b0ee8309a5efb6855dcb	unavailable	2017-10-10 10:50:03	2017-10-14 18:35:57	NaT	NaT	2017-10-23

	geolocation_zip_code_prefix	geolocation_lat	geolocation_lng	geolocation_city	geolocation_state	regiao
0	01037	-23.545621	-46.639292	sao paulo	SP	Sudeste
1	01046	-23.546081	-46.644820	sao paulo	SP	Sudeste
2	01046	-23.546129	-46.642951	sao paulo	SP	Sudeste
3	01041	-23.544392	-46.639499	sao paulo	SP	Sudeste
4	01035	-23.541578	-46.641607	sao paulo	SP	Sudeste

	order_id	price
0	00010242fe8c5a6d1ba2dd792cb16214	58.90
1	00018f77f2f0320c557190d7a144bdd3	239.90
2	000229ec398224ef6ca0657da4fc703e	199.00
3	00024acbcdf0a6daa1e931b038114c75	12.99
4	00042b26cf59d7ce69dfabb4e55b4fd9	199.90

	order_approved_at	price
0	2016-09	134.97
1	2016-10	40941.30
2	2016-12	10.90
3	2017-01	106888.10
4	2017-02	234163.38

	Mes-2017	Faturamento-2017
0	Janeiro	106888.10
1	Fevereiro	234163.38
2	Março	355372.21
3	Abril	338207.85
4	Maio	490517.21

	order_id	order_approved_at	weekday
0	e481f51cbdc54678b7cc49136f2d6af7	2017	0
1	53cdb2fc8bc7dce0b6741e2150273451	2018	3
2	47770eb9100c2d0c44946d9cf07ec65d	2018	2
3	949d5b44dbf5de918fe9c16f97b45f8a	2017	5
4	ad21c59c0840e6cb83a9ceb5573f8159	2018	1

	year	weekday	orders
0	2016	0	40
1	2016	1	44
2	2016	2	11
3	2016	3	97
4	2016	4	48
5	2016	5	23
6	2016	6	9
7	2017	0	5285
8	2017	1	8096
9	2017	2	6971
10	2017	3	6399
11	2017	4	6582
12	2017	5	5709
13	2017	6	4312
14	2018	0	7290
15	2018	1	10504
16	2018	2	8348
17	2018	3	8549
18	2018	4	7588
19	2018	5	6104
20	2018	6	4452

year	weekday	2016	2017	2018
0	Domingo	40	5285	7290
1	Segunda	44	8096	10504
2	Terça	11	6971	8348
3	Quarta	97	6399	8549
4	Quinta	48	6582	7588

	order_id	order_approved_at	Data	Ano	weekday
0	e481f51cbdc54678b7cc49136f2d6af7	2017-10-02 11:07:15	2017-10-02	2017	0
1	53cdb2fc8bc7dce0b6741e2150273451	2018-07-26 03:24:27	2018-07-26	2018	3
2	47770eb9100c2d0c44946d9cf07ec65d	2018-08-08 08:55:23	2018-08-08	2018	2
3	949d5b44dbf5de918fe9c16f97b45f8a	2017-11-18 19:45:59	2017-11-18	2017	5
4	ad21c59c0840e6cb83a9ceb5573f8159	2018-02-13 22:20:29	2018-02-13	2018	1

	purchase_part_day	order_id	percent
0	Madrugada	4593	4.76%
1	Manhã	21591	22.38%
2	Noite	33102	34.32%
3	Tarde	37169	38.54%

	order_id	payment_sequential	payment_type	payment_installments	payment_value
0	b81ef226f3fe1789b1e8b2acac839d17	1	credit_card	8	99.33
1	a9810da82917af2d9aefd1278f1dcfa0	1	credit_card	1	24.39
2	25e8ea4e93396b6fa0d3dd708e76c1bd	1	credit_card	1	65.71
3	ba78997921bbcdc1373bb41e913ab953	1	credit_card	8	107.78
4	42fdf880ba16b47b59251dd489d4441a	1	credit_card	2	128.45

	order_id	customer_id	order_status	order_purchase_timestamp	order_approved_at	order_delivered_carrier_date	order_delivered_customer_date	order_estimated_delivery_date
0	e481f51cbdc54678b7cc49136f2d6af7	9ef432eb6251297304e76186b10a928d	delivered	2017-10	2017-10	2017-10	2017-10	2017-10
1	53cdb2fc8bc7dce0b6741e2150273451	b0830fb4747a6c6d20dea0b8c802d7ef	delivered	2018-07	2018-07	2018-07	2018-08	2018-08
2	47770eb9100c2d0c44946d9cf07ec65d	41ce2a54c0b03bf3443c3d931a367089	delivered	2018-08	2018-08	2018-08	2018-08	2018-09
3	949d5b44dbf5de918fe9c16f97b45f8a	f88197465ea7920adcdbec7375364d82	delivered	2017-11	2017-11	2017-11	2017-12	2017-12
4	ad21c59c0840e6cb83a9ceb5573f8159	8ab97904e6daea8866dbdbc4fb7aad2c	delivered	2018-02	2018-02	2018-02	2018-02	2018-02

	order_id	customer_id	order_status	order_purchase_timestamp	order_approved_at	order_delivered_carrier_date	order_delivered_customer_date	order_estimated_delivery_date	price
0	e481f51cbdc54678b7cc49136f2d6af7	9ef432eb6251297304e76186b10a928d	delivered	2017-10	2017-10	2017-10	2017-10	2017-10	29.99
1	53cdb2fc8bc7dce0b6741e2150273451	b0830fb4747a6c6d20dea0b8c802d7ef	delivered	2018-07	2018-07	2018-07	2018-08	2018-08	118.70
2	47770eb9100c2d0c44946d9cf07ec65d	41ce2a54c0b03bf3443c3d931a367089	delivered	2018-08	2018-08	2018-08	2018-08	2018-09	159.90
3	949d5b44dbf5de918fe9c16f97b45f8a	f88197465ea7920adcdbec7375364d82	delivered	2017-11	2017-11	2017-11	2017-12	2017-12	45.00
4	ad21c59c0840e6cb83a9ceb5573f8159	8ab97904e6daea8866dbdbc4fb7aad2c	delivered	2018-02	2018-02	2018-02	2018-02	2018-02	19.90

	Mes-2017	Faturamento-2017	Mes-2018	Faturamento-2018	crescimento-2017-para-2018
0	Janeiro	106888.10	Janeiro	917667.20	+858.53%
1	Fevereiro	234163.38	Fevereiro	819385.26	+349.92%
2	Março	355372.21	Março	963604.25	+271.15%
3	Abril	338207.85	Abril	953526.94	+281.94%
4	Maio	490517.21	Maio	999867.28	+203.84%
5	Junho	425825.55	Junho	859396.90	+201.82%
6	Julho	476556.89	Julho	849351.12	+178.23%
7	Agosto	555061.46	Agosto	862639.54	+155.41%

year	weekday	2016	2017	2018
0	Domingo	40	103	214
1	Segunda	22	158	300
2	Terça	5	136	245
3	Quarta	32	123	251
4	Quinta	24	126	223
5	Sexta	23	109	184
6	Sabado	9	82	139

	order_id	payment_sequential	payment_installments	payment_value
payment_type
boleto	19784	19784	19784	19784
credit_card	76795	76795	76795	76795
debit_card	1529	1529	1529	1529
not_defined	3	3	3	3
voucher	5775	5775	5775	5775

	customer_state	customer_id	percent
0	SP	41746	41.98%
1	RJ	12852	12.92%
2	MG	11635	11.7%
3	RS	5466	5.5%
4	PR	5045	5.07%

	regiao	customer_id	percent	percent_number
3	Sudeste	68266	68.65%	68.65
4	Sul	14148	14.23%	14.23
1	Nordeste	9394	9.45%	9.45
0	CentroOeste	5782	5.81%	5.81
2	Norte	1851	1.86%	1.86

	customer_city	payment_value	% of Total Payments	% Cumulativa dos Pagamentos
3576	sao paulo	2826299.96	14.020701	14.020701
3136	rio de janeiro	1557926.74	7.728558	21.749259
451	belo horizonte	496991.57	2.465474	24.214734
556	brasilia	430642.95	2.136332	26.351066
1139	curitiba	328703.60	1.630632	27.981698

	customer_state	payment_value	% of Total Payments	% Cumulativa dos Pagamentos
25	SP	7553411.48	37.470942	37.470942
18	RJ	2747860.23	13.631577	51.102519
10	MG	2310633.09	11.462582	62.565101
22	RS	1135485.31	5.632912	68.198013
17	PR	1058906.29	5.253019	73.451032

	customer_state	customer_unique_id	percent
0	SP	40302	41.92%
1	RJ	12384	12.88%
2	MG	11259	11.71%
3	RS	5277	5.49%
4	PR	4882	5.08%

	order_id	order_item_id	product_id	seller_id	shipping_limit_date	price	freight_value
0	00010242fe8c5a6d1ba2dd792cb16214	1	4244733e06e7ecb4970a6e2683c13e61	48436dade18ac8b2bce089ec2a041202	2017-09-19 09:45:35	58.90	13.29
1	00018f77f2f0320c557190d7a144bdd3	1	e5f2d52b802189ee658865ca93d83a8f	dd7ddc04e1b6c2c614352b383efe2d36	2017-05-03 11:05:13	239.90	19.93
2	000229ec398224ef6ca0657da4fc703e	1	c777355d18b72b67abbeef9df44fd0fd	5b51032eddd242adc84c38acab88f23d	2018-01-18 14:48:30	199.00	17.87
3	00024acbcdf0a6daa1e931b038114c75	1	7634da152a4610f1595efa32f14722fc	9d7a1d34a5052409006425275ba1c2b4	2018-08-15 10:10:18	12.99	12.79
4	00042b26cf59d7ce69dfabb4e55b4fd9	1	ac6c3623068f30de03045865e4e10089	df560393f3a51e74553ab94004ba5c87	2017-02-13 13:57:51	199.90	18.14

	product_id	product_category_name	product_name_lenght	product_description_lenght	product_photos_qty	product_weight_g	product_length_cm	product_height_cm	product_width_cm
0	1e9e8ef04dbcff4541ed26657ea517e5	perfumaria	40.0	287.0	1.0	225.0	16.0	10.0	14.0
1	3aa071139cb16b67ca9e5dea641aaa2f	artes	44.0	276.0	1.0	1000.0	30.0	18.0	20.0
2	96bd76ec8810374ed1b65e291975717f	esporte_lazer	46.0	250.0	1.0	154.0	18.0	9.0	15.0
3	cef67bcfe19066a932b7673e239eb23d	bebes	27.0	261.0	1.0	371.0	26.0	4.0	26.0
4	9dc1a7de274444849c219cff195d0b71	utilidades_domesticas	37.0	402.0	4.0	625.0	20.0	17.0	13.0

	review_id	order_id	review_score	review_comment_title	review_comment_message	review_creation_date	review_answer_timestamp
0	7bc2406110b926393aa56f80a40eba40	73fc7af87114b39712e6da79b0a377eb	4	NaN	NaN	2018-01-18 00:00:00	2018-01-18 21:46:59
1	80e641a11e56f04c1ad469d5645fdfde	a548910a1c6147796b98fdf73dbeba33	5	NaN	NaN	2018-03-10 00:00:00	2018-03-11 03:05:13
2	228ce5500dc1d8e020d8d1322874b6f0	f9e4b658b201a9f2ecdecbb34bed034b	5	NaN	NaN	2018-02-17 00:00:00	2018-02-18 14:36:24
3	e64fb393e7b32834bb789ff8bb30750e	658677c97b385a9be170737859d3511b	5	NaN	Recebi bem antes do prazo estipulado.	2017-04-21 00:00:00	2017-04-21 22:02:06
4	f7c4243c7fe1938f181bec41a392bdeb	8e6bfb81e283fa7e4f11123a3fb894f1	5	NaN	Parabéns lojas lannister adorei comprar pela I...	2018-03-01 00:00:00	2018-03-02 10:26:53

	review_score	order_id
product_id
00066f42aeeb9f3007548bb9d3f33c38	5.0	1
86743ff92eee3d16b7df59cddd583b8c	5.0	1
868ceb027ab706a4dee42e2220006b85	5.0	1
868969d3a93aeeab7bfcd4fc3d3d65de	5.0	1
868766ed2172644fdd977d6bd395a107	5.0	1

Seleção da Database¶

Database:¶

Descrição:¶

Problema de Negócio:¶

1. Carregamento dos Dados¶

1.1 Adicionando Regiões de cada Estado¶

2. Analise Exploratória¶

2.1 Evolução das Vendas Mensais¶

2.2 Evolução Faturamento Mensal¶

2.3 Faturamento Mensal 2017 Vs 2018¶

2.4 Taxa Percentual de Crescimento Mensal¶

2.5 Existe diferença significativa de venda para cada um dos meses?¶

2.6 Existe diferença significativa de venda para cada um dos dias da semana?¶

2.6.1 Total de Vendas por Dia da Semana¶

2.6.2 Total de Vendas por Dia da Semana por Ano¶

2.6.3 Média de Vendas por Dia da Semana por Ano¶

2.7 Existe diferença significativa de venda para cada um dos periodos do dia?¶

2.7.1 Total de Vendas por Período do Dia¶

2.8 Existe diferença significativa de venda para cada um dos metodos de pagamento?¶

2.8.1 Métodos de Pagamentos¶

2.9 Existe diferença significativa de venda por região?¶

2.9.1 Total de Vendas por Estado no Brasil¶

2.9.2 Total de Vendas por Região no Brasil¶

2.9.3 Top 10 Cidades com Maior Numero de Vendas¶

2.9.4 % Contribuição Cumulativa das Vendas nas Cidades¶

2.9.5 Total de Vendas por Estado em Porcentagem¶

2.9.6 % Contribuição Cumulativa das Vendas nos Estados¶

2.10 Existe diferença significativa de numero de clientes por região?¶

2.10.1 Total de Clientes por Estado¶

2.11 Quais produtos e vendas possuem maior numero de vendas?¶

2.11.1 Top 10 Categorias Mais Vendidas¶

2.11.2 % Contribuição Cumulativa das Vendas nas Categorias¶

2.12 As notas de review do produto possuem impacto no número de vendas?¶

2.12.1 Vendas por Score (Score Médio por Produto)¶

3. WordCloud de Reviews¶

4. A localização impacta no tempo de entrega?¶

5. O tempo de entrega impacta no review do produto?¶

5.1 Diferença de Entrega entre Previsto e Real em Dias¶

6. Quais categorias são mais consumidas por região?¶

6.1 Categorias Mais Vendidas por Estado¶

6.2 Categorias Mais Vendidas nos Estados de Maior Demanda¶

6.3 Categorias Mais Vendidas entre 2016 e 2018¶

7. Clientes Ativos por Mês¶

7.1 Taxa de Retenção de Clientes¶

7.2 Faturamento de Clientes Antigos Vs Clientes Novos¶

8. Analise Geoespacial no Brasil¶

8.1 Preparando os dados para construção dos gráficos¶

8.2 Pedidos Realizados pelo Brasil¶

8.3 Regiões com maior faturamento no Brasil¶

8.4 Regiões com maior taxa de transporte em relação ao pedido¶

8.5 Tempo Medio de Entrega¶

8.6 Review Medio Por Região¶

O resto é história¶

	score_medio	numero_pedidos
product_id
aca2eb7d00ea1a7b8ebd4e68314663af	4.0	527
99a4788cb24856965c36a24e339b6058	4.0	491
422879e10f46682990de24d770e7f83d	4.0	487
389d119b48cf3043d311335e499d9c6b	4.0	392
368c6c730842d78016ad823897a372db	4.0	391

	score_medio	numero_pedidos	percent
0	1.0	3798	3.35%
1	2.0	3723	3.29%
2	3.0	12364	10.91%
3	4.0	62359	55.03%
4	5.0	31078	27.42%

	review_comment_message
3	Recebi bem antes do prazo estipulado.
4	Parabéns lojas lannister adorei comprar pela I...
9	aparelho eficiente. no site a marca do aparelh...
12	Mas um pouco ,travando...pelo valor ta Boa.\r\n
15	Vendedor confiável, produto ok e entrega antes...
...	...
99983	Entregou dentro do prazo. O produto chegou em ...
99990	O produto não foi enviado com NF, não existe v...
99996	Excelente mochila, entrega super rápida. Super...
99998	Solicitei a compra de uma capa de retrovisor c...
99999	meu produto chegou e ja tenho que devolver, po...

	review_comment_message
3	recebi bem ante prazo estipulado
4	parabns lojas lannister adorei comprar interne...
9	aparelho eficiente site marca aparelho impress...
12	pouco travando valor ta boa
15	vendedor confivel produto ok entrega ante prazo

	order_id	product_id	review_score	order_delivered_carrier_date	order_delivered_customer_date	order_estimated_delivery_date
0	00010242fe8c5a6d1ba2dd792cb16214	4244733e06e7ecb4970a6e2683c13e61	5	2017-09-19 18:34:16	2017-09-20 23:43:48	2017-09-29 00:00:00
1	00018f77f2f0320c557190d7a144bdd3	e5f2d52b802189ee658865ca93d83a8f	4	2017-05-04 14:35:00	2017-05-12 16:04:24	2017-05-15 00:00:00
2	000229ec398224ef6ca0657da4fc703e	c777355d18b72b67abbeef9df44fd0fd	5	2018-01-16 12:36:48	2018-01-22 13:19:16	2018-02-05 00:00:00
3	00024acbcdf0a6daa1e931b038114c75	7634da152a4610f1595efa32f14722fc	4	2018-08-10 13:28:00	2018-08-14 13:32:39	2018-08-20 00:00:00
4	00042b26cf59d7ce69dfabb4e55b4fd9	ac6c3623068f30de03045865e4e10089	5	2017-02-16 09:46:09	2017-03-01 16:42:31	2017-03-17 00:00:00

	order_id	product_id	review_score	order_delivered_carrier_date	order_delivered_customer_date	order_estimated_delivery_date
0	00010242fe8c5a6d1ba2dd792cb16214	4244733e06e7ecb4970a6e2683c13e61	5	2017-09-19	2017-09-20	2017-09-29
1	00018f77f2f0320c557190d7a144bdd3	e5f2d52b802189ee658865ca93d83a8f	4	2017-05-04	2017-05-12	2017-05-15
2	000229ec398224ef6ca0657da4fc703e	c777355d18b72b67abbeef9df44fd0fd	5	2018-01-16	2018-01-22	2018-02-05
3	00024acbcdf0a6daa1e931b038114c75	7634da152a4610f1595efa32f14722fc	4	2018-08-10	2018-08-14	2018-08-20
4	00042b26cf59d7ce69dfabb4e55b4fd9	ac6c3623068f30de03045865e4e10089	5	2017-02-16	2017-03-01	2017-03-17

	order_id	product_id	review_score	order_delivered_carrier_date	order_delivered_customer_date	order_estimated_delivery_date
8	0005a1a1728c9d785b8e2b08b904576c	310ae3c140ff94b03219ad0adc3c778f	1	2018-03-28	2018-03-29	2018-03-29
11	00063b381e2406b52ad429470734ebd5	f177554ea93259a5b282f24e33f65ab6	5	2018-07-30	2018-08-07	2018-08-07
101	00378c6c981f234634c0b9d6128df6dd	38fa750a3a3b3204f169c86a3284d387	4	2018-02-05	2018-02-26	2018-02-26
231	0097f0545a302aafa32782f1734ff71c	b6397895a17ce86decd60b898b459796	1	2018-03-06	2018-03-21	2018-03-21
232	0097f0545a302aafa32782f1734ff71c	636598095d69a5718e67d2c9a3c7dde6	1	2018-03-06	2018-03-21	2018-03-21

	customer_id	customer_state	order_id	product_category_name
0	06b8999e2fba1a1fbc88172c00ba8bc7	SP	00e7ee1b050b8499577073aeb2a297a1	moveis_escritorio
1	18955e83d337fd6b2def6b18a428ac77	SP	29150127e6685892b6eab3eec79f59c7	utilidades_domesticas
2	4e7b3e00288586ebd08712fdd0374a03	SP	b2059ed67ce144a36e2aa97d2c9e9ad2	moveis_escritorio
3	b2b6027bc5c5109e529d4dc6358b12c3	SP	951670f92359f4fe4a63112aa7306eba	moveis_escritorio
4	4f2d8ab171c80ec8364f7c12e35b23ad	SP	6b7d50bd145f6fc7f33cebabd7e49d0f	casa_conforto

	index	product_category_name
0	cama_mesa_banho	5235
1	beleza_saude	4204
2	esporte_lazer	3667
3	moveis_decoracao	3531
4	utilidades_domesticas	3265

	order_purchase_timestamp	product_category_name
34053	2016-09-15	beleza_saude
34052	2016-09-15	beleza_saude
34051	2016-09-15	beleza_saude
108516	2016-10-03	esporte_lazer
103314	2016-10-03	relogios_presentes